intelligent compaction control of highway embankment soil
TRANSCRIPT
Report No. K-TRAN: KSU-06-7FINAL REPORT
IntellIgent CompaCtIon Control of hIghway embankment soIl In kansas
Farhana RahmanMustaque Hossain, Ph.D., P.E.Stefan Romanoschi, Ph.D., P.E.*Kansas State University*currently affiliated with The University of Texas at Arlington
March 2008
A COOPERATIvE TRANSPORTATION RESEARCH PROgRAMbETwEEN:
KANSAS DEPARTMENT OF TRANSPORTATIONKANSAS STATE UNIvERSITyTHE UNIvERSITy OF KANSAS
1 Report No.K-TRAN: KSU-06-7
2 government accession no. 3 recipient Catalog no.
4 title and subtitleIntelligent Compaction Control of Highway Embankment Soil in Kansas
5 report DateMarch 2008
6 performing organization Code
7 author(s)Farhana Rahman, Mustaque Hossain, Ph.D., P.E., Stefan Romanoschi, Ph.D., P.E.**currently affiliated with The University of Texas at Arlington
8 performing organization report no.
9 performing organization name and addressDepartment of Civil EngineeringKansas State University2118 Fiedler HallManhattan, Kansas 66506
10 work Unit no. (traIs)
11 Contract or grant no. C1571
12 sponsoring agency name and addressKansas Department of TransportationBureau of Materials and Research700 Sw Harrison StreetTopeka, Kansas 66603-3745
13 type of report and period CoveredFinal ReportAugust 2005-Fall 2007
14 sponsoring agency Code RE-0408-01
15 supplementary notesFor more information write to address in block 9.
16 abstract
Mechanistic pavement design procedures based on elastic layer theory require characterization of pavement layer materials including subgrade soil. This paper discusses the subgrade stiffness measurements obtained from a new compaction roller for compaction control on highway embankment projects in Kansas. Three test sections were compacted using a single, smooth steel drum intelligent compaction (IC) roller that compacts and simultaneously measures stiffness values of the compacted soil. Traditional compaction control measurements such as, density, in-situ moisture content, soil stiffness measurements using soil stiffness gage, surface deflection tests using the Light Falling Weight Deflectometer (LFWD) and Falling Weight Deflectometer (FWD), and penetration tests using a Dynamic Cone Penetrometer (DCP), were also done. The results show that the IC roller was able to identify the locations of lower soil stiffness in the spatial direction. Thus the IC roller can be used in proof rolling. IC roller stiffness showed sensitivity to the field moisture content indicating that moisture control during compaction is critical. No universal correlation was observed among the IC roller stiffness, soil gage stiffness, backcalculated subgrade moduli from the LFWD and FWD deflection data, and the California Bearing Ratio (CBR) obtained from DCP tests. The discrepancy seems to arise from the fact that different equipment were capturing response from different volumes of soil on the same test section. Analysis using the newly released Mechanistic-Empirical Pavement Design Guide (M-EPDG) shows that pavement rutting, roughness and asphalt base thickness are significantly influenced by the subgrade strength. “Target” modulus for compaction quality control can also be obtained by this analysis.
17 key wordsIntelligent compaction, falling weight deflectometer (FWD), light falling weight deflectometer (LFWD), stiffness gage, nuclear gage, dynamic cone penetrometer (DCP), moisture content, soil stiffness
18 Distribution statementNo restrictions. This document is available to the public through the National Technical Information Service,Springfield, Virginia 22161
19 security Classification (of this report)
Unclassified
20 security Classification (of this page) Unclassified
21 no. of pages 122
22 price
Form DOT F 1700.7 (8-72)
IntellIgent CompaCtIon Control of hIghway embankment soIl In kansas
final report
Prepared by
Farhana RahmanMustaque Hossain, Ph.D., P.E.
Stefan Romanoschi, Ph.D., P.E.*
Kansas State University2118 Fiedler Hall
Manhattan, Kansas 66506
*currently affiliated with The University of Texas at Arlington
A Report on Research Sponsored By
THE KANSAS DEPARTMENT OF TRANSPORTATIONTOPEKA, KANSAS
March 2008
© Copyright 2008, kansas Department of transportation
ii
prefaCe
The Kansas Department of Transportation’s (KDOT) Kansas Transportation Research and New-Developments (K-TRAN) Research Program funded this research project. It is an ongoing, cooperative and comprehensive research program addressing transportation needs of the state of Kansas utilizing academic and research resources from KDOT, Kansas State University and The University of Kansas. Transportation professionals in KDOT and the universities jointly develop the projects included in the research program.
notICe
The authors and the state of Kansas do not endorse products or manufacturers. Trade and manufacturers’ names appear herein solely because they are considered essential to the object of this report.
This information is available in alternative accessible formats. To obtain an alternative format, contact the Office of Transportation Information, Kansas Department of Transportation, 700 SW Harrison, Topeka, Kansas 66603-3745 or phone (785) 296-3585 (Voice) (TDD).
DIsClaImer
The contents of this report reflect the views of the authors who are responsible for the facts and accuracy of the data presented herein. The contents do not necessarily reflect the views or the policies of the state of Kansas. This report does not constitute a standard, specification or regulation.
iii
ABSTRACT
Mechanistic pavement design procedures based on elastic layer theory require
characterization of pavement layer materials including subgrade soil. This report
discusses the subgrade stiffness measurements obtained from a new compaction roller
for compaction control on highway embankment projects in Kansas. Three test sections
were compacted using a single, smooth steel drum intelligent compaction (IC) roller that
compacts and simultaneously measures stiffness values of the compacted soil.
Traditional compaction control measurements such as, density, in-situ moisture content,
soil stiffness measurements using soil stiffness gage, surface deflection tests using the
Light Falling Weight Deflectometer (LFWD) and Falling Weight Deflectometer (FWD),
and penetration tests using a Dynamic Cone Penetrometer (DCP), were also done. The
results show that the IC roller was able to identify the locations of lower soil stiffness in
the spatial direction. Thus the IC roller can be used in proof rolling. IC roller stiffness
showed sensitivity to the field moisture content indicating that moisture control during
compaction is critical. No universal correlation was observed among the IC roller
stiffness, soil gage stiffness, backcalculated subgrade moduli from the LFWD and FWD
deflection data, and the California Bearing Ratio (CBR) obtained from DCP tests. The
discrepancy seems to arise from the fact that different equipments were capturing
response from different volumes of soil on the same test section. Analysis using the
newly released Mechanistic-Empirical Pavement Design Guide (M-EPDG) shows that
pavement rutting, roughness and base thickness are significantly influenced by the
subgrade strength. “Target” modulus for compaction quality control can also be
obtained by the analysis.
iv
ACKNOWLEDGEMENTS
The authors wish to acknowledge the financial support provide by the Kansas
Department of Transportation under its Kansas Transportation and New Developments
(K-TRAN) program. Contribution of Ms. Mbaki Onyongo, Mr. Christian Dumitru, and Mr.
James Mulandi of Kansas State University is gratefully acknowledged. Participation of
Bomag Corp. in this study is also gratefully acknowledged. The help of KDOT area
personnel and contractors is also acknowledged. Dr. James Neil of the Statistics
Department at Kansas State University served as the statistical consultant to this
project. His contribution is duly acknowledged.
v
TABLE OF CONTENTS
ABSTRACT ..................................................................................................................... iii
ACKNOWLEDGEMENTS ...............................................................................................iv
Table of Contents ............................................................................................................ v
List of Figures ................................................................................................................ viii
List of Tables ................................................................................................................... x
CHAPTER 1 - INTRODUCTION ..................................................................................... 1
1.1 General .............................................................................................................. 1
1.2 Convention In-Situ Quality Control (Spot Tests) ................................................ 2
1.3 Relative Compaction Testing Methods .............................................................. 4
1.4 Absolute Compaction Testing Method ............................................................... 5
1.5 Intelligent Compaction Method .......................................................................... 5
1.6 Problem Statements .......................................................................................... 5
1.7 Objectives .......................................................................................................... 8
1.8 Report Organization ........................................................................................... 8
CHAPTER 2 - LITERATURE REVIEW ............................................................................ 9
2.1 Soil Compaction ................................................................................................. 9
2.1.1 Compaction Principles and Measurements ...................................................... 10
2.1.2 Factors Affecting Field Compaction ................................................................. 12
2.1.3 Field Tests for Compaction Control.................................................................. 12
2.1.4 Compaction Quality Control in the Field ........................................................... 13
2.2 Intelligent Compaction Control ......................................................................... 14
2.2.1 Historical Background ...................................................................................... 15
2.2.2 Background Mechanics of Intelligent Compaction ........................................... 16
2.2.3 Compaction Measurement and Documentation Systems of IC Roller ............. 22
2.2.3.1 Measurement System ...................................................................................... 22
2.2.3.2 Measurement Principle .................................................................................... 24
CHAPTER 3 - FIELD TESTING AND DATA COLLECTION ......................................... 27
3.1 General ............................................................................................................ 27
3.2 Description of Test Section .............................................................................. 27
vi
3.2.1 Laboratory Testing ........................................................................................... 28
3.3 IC Roller Description ........................................................................................ 29
3.4 In-situ Testing and Data Collection .................................................................. 30
3.4.1 Density and Moisture Measurements ............................................................... 31
3.4.2 Stiffness, Deflection, and Cone Penetration Tests ........................................... 31
3.4.2.1 Geogage .......................................................................................................... 32
3.4.2.2 Falling Weight Deflectometer (FWD) ............................................................... 33
3.4.2.3 Light Falling Weight Deflectometer (LFWD) ..................................................... 35
3.4.2.4 Dynamic Cone Penetrometer (DCP) ................................................................ 36
CHAPTER 4 - RESULTS AND DISCUSSION ............................................................... 39
4.1 General ............................................................................................................ 39
4.2 BVC Stiffness on US-56 “Proof”, “Growth” and I-70 “Growth” Sections ........... 40
4.3 Variation of IC Roller Stiffness with Moisture Content...................................... 43
4.4 Variation of IC Stiffness with Compaction Level .............................................. 48
4.5 Variation of IC Roller, Geogage, LFWD, and FWD Stiffness ........................... 51
4.6 Statistical Analysis of Test Results .................................................................. 55
4.6.1 Correlation Matrix............................................................................................. 57
4.6.2 Reasons for Poor Correlation among Different Measures of Stiffness and
Moduli .............................................................................................................. 59
CHAPTER 5 - MECHANISTIC-EMPIRICAL PAVEMENT DESIGN (M-E PDG)
ANALYSIS ....................................................................................................... 61
5.1 General ............................................................................................................ 61
5.2 M-E PDG Design Approach ............................................................................. 61
5.3 Overview of the Design Process of M-EPDG ................................................... 62
5.4 Test Sections ................................................................................................... 65
5.5 Design Inputs of M-EPDG (Version 1.0) Analysis ............................................ 67
5.5.1 General Inputs ................................................................................................. 69
5.5.2 Design Traffic Inputs ........................................................................................ 71
5.5.3 Climate Inputs .................................................................................................. 72
5.5.4 Structural Inputs ............................................................................................... 73
5.5.4.1 US-56 Full Depth Asphalt Concrete Pavement Section ................................... 73
vii
5.5.4.2 I-70 Full Depth Asphalt Concrete and PCC Pavement Sections ...................... 73
5.6 Prediction on Distresses from M-EPDG Analysis ............................................ 79
5.6.1 International Roughness Index (IRI) ................................................................ 79
5.6.2 Total Deformation ............................................................................................ 81
5.6.3 Effect of Subgrade Modulus on Design PCC Slab Thickness .......................... 85
5.7 Determination of Target Modulus by M-EPDG Analysis .................................. 87
5.8 Estimation of the Regression Constants for M-E Pavement Design ................ 90
CHAPTER 6 - CONCLUSIONS AND RECOMMENDATIONS ...................................... 96
6.1 Conclusions ..................................................................................................... 96
6.2 Recommendations ........................................................................................... 98
References .................................................................................................................... 99
APPENDIX A ............................................................................................................... 105
SAS Nonlinear Regression Analysis ........................................................................... 105
Input and Output Files ................................................................................................. 105
viii
LIST OF FIGURES
FIGURE 1.1: Dynatest Model 8000 FWD device used in spot testing on US-56,
Hugoton, Kansas ....................................................................................................... 3
FIGURE 1.2: Omegameter and Terrameter (2) ............................................................... 4
FIGURE 1.3: Comparison between Conventional roller and IC roller, (6) ....................... 7
FIGURE 2.1: Relation of dry unit weight to the moisture content .................................. 10
FIGURE 2.2: Application of Intelligent Compaction in US (7) ........................................ 16
FIGURE 2.3: Theoretical, Lumped parameter model of interaction between a vibratory
roller and underneath material, (15) ......................................................................... 18
FIGURE 2.4: Loading loop due to applied force on the roller drum, (7) ........................ 19
FIGURE 2.5: Soil reactions vs. roller amplitudes (after 7) ............................................. 20
FIGURE 2.6: Hertz (1895), and Lundberg (1939) solution for modulus (7) ................... 21
FIGURE 2.7: Relationship between stiffness and modulus (6) ...................................... 22
FIGURE 2.8: BOMAG measurement and documentation system, (2)........................... 23
FIGURE 2.9: BOMAG Intelligent Compaction System, (2) ............................................ 24
FIGURE 2.10: Direction of applied vibration to optimize compaction, (7) ...................... 26
FIGURE 3.1: Moisture-Density curves from Standard Proctor Test .............................. 29
FIGURE 3.2: BVC roller compaction on a test section on US-56 near Hugoton, KS ..... 30
FIGURE 3.3: In-situ moisture and density measurement using nuclear gage on I-70 ... 31
FIGURE 3.4: Stiffness measurements by Geogage, on US-56 section ........................ 32
FIGURE 3.5: Deflection measurements on US-56 section by FWD .............................. 33
FIGURE 3.6: Normalized deflection basins on US-56 section (station 15+00) by
EVERCALC 5.0........................................................................................................ 34
FIGURE 3.7: Deflection measurements by LFWD on I-70 section ................................ 35
FIGURE 3.8: DCP testing on I-70 test section in Kansas .............................................. 37
FIGURE 4.1: Stiffness developed and measured by IC roller on I-70 and US-56 ......... 40
FIGURE 4.2: Spatial variation of IC roller stiffness on (a) US-56 “proof”, (b) US-56
“growth” and (c) I-70 “growth” .................................................................................. 42
FIGURE 4.3: Variation of IC roller stiffness with in-situ moisture content on (a) US-56
“proof”, (b) US-56 “growth” and (c) I-70 “growth” section. ........................................ 45
ix
FIGURE 4.4: Variation of IC roller stiffness with moisture differential on (a) US-56
“proof”, (b) US-56 “growth” and (c) I-70 “growth” section. ........................................ 47
FIGURE 4.5: Relationship between IC roller stiffness and percent compaction for (a)
US-56 “proof”, (b) US-56 “growth” and (c) I-70 “growth” sections ............................ 50
FIGURE 4.6: Variation of IC roller, FWD, LFWD and Geogage stiffness on (a) US-56
“proof”, (b) US-56 “growth” and (c) I-70 “growth” sections ....................................... 54
FIGURE 4.7: Volume of influence space in subgrade soil ............................................. 60
FIGURE 5.1: Design framework of M-EPDG 2000, (1) ................................................. 62
FIGURE 5.2: M-EPDG design process for flexible pavement, (1) ................................. 64
FIGURE 5.3: M-EPDG design process for JPCP, (1) .................................................... 64
FIGURE 5.4: Full Depth Flexible Pavement (US-56, I-70 Section) and PCCP (I-70)
layered system ......................................................................................................... 66
FIGURE 5.5: Predicted total deformation on (a) US-56 “proof” and (b) US-56 “growth”
and (c) I-70 section .................................................................................................. 82
FIGURE 5.6: Effect of subgrade strength on total deformation at different base
thickness .................................................................................................................. 84
FIGURE 5.7: Evaluation of (a) IRI and (b) % slabs cracked at different slab thickness
based on subgrade modulus .................................................................................... 86
FIGURE 5.8: “Target” subgrade modulus @ 90% reliability on (a) I-70 and (b) US-56 . 88
FIGURE 5.9: Fitness of Mr data in SAS Nonlinear regression (a) (No. of Obs.=15), (b)
(No. of Obs.=6) ........................................................................................................ 95
x
LIST OF TABLES
TABLE 3.1: Test Section Soil Characteristics ............................................................... 28
TABLE 4.1: Applied Vertical Stresses and Moisture Content During IC Compaction
and Testing ............................................................................................................... 52
TABLE 4.2: Statistical Summary of Stiffness and Moduli Results ................................. 56
TABLE 4.3: Correlations among IC and Geogage Stiffness, LFWD and FWD Moduli,
and CBR on US-56 “Proof” Section ........................................................................... 57
TABLE 4.4: Correlations among BVC, Geogage, FWD Moduli, and CBR on US-56
“Growth” Section ....................................................................................................... 58
TABLE 4.5: Correlations among BVC, Geogage, LFWD and FWD Moduli, and CBR
on I-70 “Growth” Section ........................................................................................... 58
TABLE 5.1: Project Details of the Test Sections ........................................................... 66
TABLE 5.2: Performance Criteria for JPC Pavement and Full Depth Asphalt
Concrete Pavement .................................................................................................. 71
TABLE 5.3: Layer and Material Inputs for Flexible Pavement Analysis on I-70 and
US-56 ........................................................................................................................ 75
TABLE 5.4: Input Parameters for M-EPDG Rigid Pavement (JPCP) Analysis on I-70
Test Section .............................................................................................................. 76
TABLE 5.5: Subgrade Properties for Flexible/Rigid Pavement Analysis of I-70 and
US-56 Test Sections ................................................................................................. 78
TABLE 5.6: IRI Evaluations by M-EPDG Analysis ........................................................ 80
TABLE 5.7: “Target” Modulus and Measured/Calculated Modulus on US-56 and I-70
Sections .................................................................................................................... 89
TABLE 5.8: Laboratory Resilient Modulus Test Results on US-56 and I-70 Sections ... 92
TABLE 5.9: Estimated Regression Coefficient k1, k2 and k3 on US-56 Section ............. 93
TABLE 5.10: Estimated Regression Coefficient k1, k2 and k3 on I-70 Section ............... 94
1
CHAPTER 1 - INTRODUCTION
1.1 General
During past four decades, pavements have been designed mostly using
empirical design procedures. However, design of pavements using new tools requires
detailed inputs on material response and damage properties. Example of such a tool is
the newly developed Mechanistic-Empirical Pavement Design Guide (M-EPDG) of
National Cooperative Highway Research Program (NCHRP) (1). Pavement
performance is highly influenced by foundation layer modulus, strength, and
permeability as well as by the ease and permanency of compaction. Subgrade
compaction increases strength, decreases permeability, and reduces undesirable
settlement. Current compaction quality control methods are fully based on the results of
the laboratory compaction tests. The in-situ dry density of the soil is measured after
compaction and compared with the laboratory maximum dry density. A number of
methods such as, sand cone, rubber balloon and nuclear gage, etc. are used to
measure the in-situ density. The in-situ moisture is also measured by the nuclear gage
and other measurement methods and controlled.
During a scan tour in Europe, the Federal Highway Administration (FHWA)
identified some road building technologies that are implementable in the United States.
The technologies are two lift construction, design feature catalog, high quality
foundation construction, greater attention to mix design components, geotextile layer,
and exposed aggregate surface. A new compaction technique named intelligent
compaction came under the high quality foundation construction technology. Intelligent
compaction was first introduced in some European countries for road and embankment
2
construction in early 1970s. The aim of this new method was to provide higher quality
road and embankment construction through a high quality control and interactive
assurance system from the very beginning of the roadway construction. This report will
describe a field testing using an intelligent compaction roller on two highway
embankment projects in Kansas as well as the associated research results.
In the United States, the soil and rock fill materials are compacted using
conventional static or vibratory rollers. The highway embankment of cohesionless soil is
mainly compacted with a vibratory roller with parallel strips passes from edge to edge or
with some overlapping. Each strip is compacted by a fixed number of passes with a
constant roller speed, vibration frequency and amplitude. However, constant speed,
frequency and amplitude do not necessarily lead to a homogeneous compaction level
due to variation in subgrade material properties, in-situ moisture content of the
compacted layer, and the stiffness of the underlying layer soil. Instead this process
often leads to a certain part of the area insufficiently compacted, some portion over-
compacted, and the rest sufficiently compacted.
1.2 Convention In-Situ Quality Control (Spot Tests)
The conventional in-situ quality control of compaction involves some spot tests
such as static plate load test, soil stiffness gage for stiffness measurement, falling
weight deflectometer (FWD) for modulus testing (Figure 1.1), the nuclear gage for
density and in-situ moisture testing, rubber balloon and the sand replacement tests for
density measurement. These methods are standardized and some are widely used for
in-situ compaction control. However, it should be noted that the zone of influence within
the compacted subgrade layer for each test method is different. For example, the
3
modulus value obtained from plate load test covers the depth that is 1.5 times the plate
diameter (ranges from 0.3 to 0.6 m). The soil stiffness gage (Geogage) provides
subgrade stiffness for a depth of 0.5 ft (0.15 m). Again, the measured modulus value
may differ significantly with depth due to heterogeneity of the natural soil. In addition,
the moisture content will affect the average density value of the compacted subgrade.
FIGURE 1.1: Dynatest Model 8000 FWD device used in spot testing
on US-56, Hugoton, Kansas
4
1.3 Relative Compaction Testing Methods
In this method, index values are compared for two successive passes of the
compaction roller. The method does not provide any absolute value for percent
compaction, stiffness or density achieved. These systems are available as an
attachment for any compaction roller and are called “Compactometer”. A Swedish
company, Geodynamic manufactures and sells this type of equipment.
This approach is also used in the “Continuous Compaction Control, CCC.” The
measurement methods are based on integrated compaction meter attached to a roller
that continuously measures the acceleration of the roller drum and continuously
calculates meter value from the acceleration signal. Examples of such measures are
compaction meter value (CMV) from GEODYNAMIK, and the Omegameter and
Terrameter (Figure 1.2) from BOMAG. The basic objective is to obtain a quality
assurance documentation which can also be used to identify the spot test locations for
calibration.
FIGURE 1.2: Omegameter and Terrameter (2)
5
1.4 Absolute Compaction Testing Method
This method provides the absolute values of compaction during operation. In the
past, the absolute compaction results were provided by the individual or independent
measuring units that were not attached to the compaction equipment. Currently, some
manufacturers are providing an attachment that gives absolute instantaneous values of
soil stiffness measures. These systems assure the operator that the proper compaction
has been achieved. A Swiss company, AMMANN has such a system attached to their
roller.
1.5 Intelligent Compaction Method
This method is a combination of the absolute measurement technology and an
automatic control method. The system includes a vibratory roller that measures the
material stiffness continuously and also has an automatic compaction control. The
system controls different compaction parameters of the roller such as, amplitude,
frequency and working speed. The stiffness measurements are made by the
instrumented roller itself. The details of intelligent compaction process will be discussed
in Chapter 2.
1.6 Problem Statements
Compaction of embankment, subgrade and base materials costs a significant
portion of state highway agency construction budgets and is critical to the performance
of highway pavements. Heterogeneity of the pavement materials, variability in the
equipment and operators, difficulty in maintaining the uniform lift thickness and
prescribed moisture content make the target compaction level difficult to achieve.
Current quality control and quality assurance testing devices, such as, nuclear gage, the
6
dynamic cone penetrometer, the geogage, the light falling weight deflectometer are
typically used to asses less than one percent of the actual compacted area (3). In
addition, each of these testing devices measures parameters unique to the device.
Other limitations of the conventional compaction and the compaction control process as
identified by FHWA (4) are: (1) Density and density related material properties can not
be measured before the compaction process is complete; (2) Density measurement
from a small number of spots may not be representative of the density of the entire lot;
(3) Conventional compaction methodology does not allow any or very little
instantaneous feedback for the project personnel; and finally (4) Overcompaction can
occur and hence reduce the density that has already been achieved in the previous
passes.
Improper compaction control of subgrade soils may result in bridge approach
settlement, rapid increase in pavement roughness, etc. The Kansas Department of
Transportation (KDOT) has placed a major emphasis on the compaction control of soils
on the “grade and pave” projects. Embankment compaction control is specified by the
KDOT based on the maximum dry density and optimum moisture content results
obtained from the standard proctor tests on typical embankment soil. Currently KDOT is
in the process of implementing quality control/quality assurance (QC/QA) specifications
for embankment compaction. A previous research project in Kansas showed that
current compaction control procedures result in highly variable stiffness values of the
finished subgrade (5).
7
The problem of variable stiffness of the pavement foundation layer has been
addressed by the European countries using Intelligent Compaction Control (ICC).
According to FHWA (4), the primary reason to consider Intelligent Compaction (IC)
technology is its stated capability to optimize and significantly improve the conventional
compaction process as shown in Figure 1.3. This improvement is in several important
areas: (1) Compaction efficiency (fewer passes); (2) Compaction quality (consistently
higher and more uniform density); (3) Continuous material stiffness outputs that can be
used in mechanistic-empirical pavement design; (4) Real time information to project
personnel during compaction; (5) Identification of situations where adequate compaction
cannot be accomplished. More details on ICC can be found elsewhere (4, 7-11). But IC
roller stiffness should be evaluated to achieve the target pavement performance for a
given design. Again, recent research has shown that the M-EPDG overestimates
pavement performance by assuming subgrade compaction to optimum moisture content
FIGURE 1.3: Comparison between Conventional roller and IC roller, (6)
8
(12). In order to ensure a successful design, the distresses on the pavement must be
predicted for in-situ subgrade modulus.
1.7 Objectives
The main objective of this study was to evaluate the soil stiffness measured by IC
roller and its correlation with other stiffness and/or modulus values obtained from
Geogage, Light Falling Weight Deflectometer (LFWD) and Falling Weight Deflectometer
(FWD) deflections, and Dynamic Cone Penetrometer (DCP) data on compacted
subgrade. Variation of IC roller measured stiffness with field moisture content and
percent compaction was also analyzed. The final objective of this study was to evaluate
the soil stiffness and/or modulus values measured by the IC roller and different non-
destructive testing (NDT) devices using the Mechanistic-Empirical Pavement Design
Guide software (version 1.0) so that a “target” modulus can be selected. Sensitivity
analysis was also performed to examine the effect of subgrade modulus on the
distresses of the whole pavement structure.
1.8 Report Organization
This report is divided into six chapters. The first chapter covers a brief
introduction to the compaction technique and its quality control, problem statement,
study objectives and the outline of the report. Chapter 2 is a review of the literature and
a detailed background of intelligent compaction. Chapter 3 describes the test section
and data collection procedure in the field. Chapter 4 presents the analysis of the test
results. Results of M-EPDG analysis for the test sections are discussed in Chapter 5.
Finally, Chapter 6 presents the conclusions and recommendations based on the present
study.
9
CHAPTER 2 - LITERATURE REVIEW
2.1 Soil Compaction
In the construction of many engineering structures such as, highway
embankment and earth dam, loose earth fills, etc. are required to be compacted to
increase the soil density and hence the load bearing capacity. Soil compaction is a
mechanical process to densify the soil by removing air mass from the void space within
the soil structures. In time, loose material would settle or compact itself naturally. By
applying various mechanical forces, the time required to get is significantly reduced.
Pavement performance is basically influenced by the subgrade strength, drainage
properties of the soil, ease of compaction and also the permanency of the compaction.
Proper soil compaction increases the bearing capacity of the soil, reduces water
seepage, swelling and shrinkage potential, protects the soil from uneven settlements
and frost damage and also provides stability of the compacted soil mass.
The degree of compaction of soil is measured in terms of dry unit weight. In
general, the dry unit weight will increase with increasing moisture content for similar
compacting effort. However, beyond a certain point, additional moisture tends to reduce
the dry unit weight as excess moisture try to displace soil particles from their compacted
position. Figure 2.1 shows the general nature of dry unit weight to moisture content for a
given soil and the compaction effort. The moisture content at which the maximum dry
unit weight is obtained is called the optimum moisture content.
10
2.1.1 Compaction Principles and Measurements
During compaction, the voids between the particles that are filled with air, water
or a combination of both are expelled by a combination of force and movement. Four
different types of forces may be present during compaction: (1) Static Pressure, (2)
Manipulation, (3) Impact Force, and (4) Vibration.
STATIC PRESSURE: In static compaction, weighted loads, applied by rollers,
produce shear stresses in the soil which cause the individual particles to slide across
each other. These particles break their natural bonds to each other and move into a
more stable position within the material. Static smooth-wheeled rollers, static
sheepsfoot (or pad-foot) and tamping foot rollers work on this principle. Four factors
influence compaction performance of static rollers. They are axle load, drum width,
drum diameter and rolling speed.
MANIPULATION: Manipulation is a compactive force that rearranges particles
into a more dense mass by a kneading process. The process is especially effective at
FIGURE 2.1: Relation of dry unit weight to the moisture content
Sp. Gravity Curves for different
Degree of Saturation, Sr
11
the surface of the lift material. The longitudinal and transverse kneading action is
essential during the compaction of heavily stratified soils such as, clay-type soils.
Manipulation helps to close the small, hairline cracks through which moisture can
penetrate and weaken the subgrade. Sheepsfoot rollers and staggered wheel, rubber
tired rollers are specifically designed to deliver this type of compactive force.
IMPACT: Impact creates a greater compaction force on the surface than an
equivalent static load. This is because a falling weight speed is converted to energy at
the instant of impact. Impact creates a pressure wave, which goes into the soil from the
surface. Impacts are usually a series of blows. Impact blows of 5 to 600 blows per
minute are considered low frequency ranges and are used on impact hammers and
hand tampers. Impact blows of high frequency (1400 to 3500 blows per minute) are
used on vibratory compactors.
VIBRATION: Vibration is the final and most complex compactive force. Vibratory
compactors produce a rapid succession of pressure waves, which spread in all
directions. The vibratory pressure waves are useful in breaking the bonds between the
particles of the material being compacted. When pressure is applied, the particles tend
to reorient themselves in a more dense (fewer voids) state.
Vibratory compaction of soil is a complex process. More than 30 different factors
influence the overall compaction effort. Vibratory compaction involves a drum which is
moving up and down (amplitude) very rapidly (frequency) and moving forward (working
speed) over a non-homogeneous material. All components influencing compaction
should be considered as a whole, not as individual entities. It is the combined
characteristics of the compactor and of the mass of soil it is attempting to compact that
12
determines the degree of compactive effort. Vibratory rollers are used mostly for
densification of the granular subgrade soil.
2.1.2 Factors Affecting Field Compaction
Field soil compaction basically depends on the soil structure and properties and
the in-situ moisture content in the field. Besides these two factors, the thickness of the
lift, the intensity of the pressure applied by the compaction equipment and the area over
which the pressure is applied, greatly affect the degree of compaction. The pressure
applied at the surface by the compaction equipment decreases with depth and hence
will decrease the degree of compaction. During compaction, the dry unit weight of soil is
also affected by the number of roller passes.
2.1.3 Field Tests for Compaction Control
Periodic field testing is done to measure two important parameters: (1) target
density and (2) target moisture content. These tests can also indicate the effectiveness
of the compaction equipment and construction method being used. The most common
field testing methods are the Nuclear Method, the Sand-Cone Method and the Water
Balloon Method.
NUCLEAR METHOD: Nuclear density meters emit radiation into the subgrade
soil being tested and count the measures (both moisture content and density). The test
is nondestructive and can be performed quickly. There are two basic methods of
measuring density in this method: (1) The direct transmission method that gives the
best accuracy, least composition error and least surface roughness error for testing over
a range of depths from two to twelve inches; and (2) The backscatter method that
eliminates the need to create an access hole in the compacted soil. However, accuracy
13
is less and composition errors are likely. This method works best in shallow depths (2 to
3 inches).
SAND CONE METHOD: The sand-cone method is a multi-step procedure which
is more time consuming than the nuclear density method, but has had proven accuracy.
It is sometimes used in conjunction with the nuclear method to verify the calibration of
the nuclear density meter.
WATER BALLON METHOD: The water balloon method is also called the
Washing Densometer Test. The first three steps of this test are to excavate a sample,
then weigh it, and dry it. These steps are same as performed in the sand-cone method.
In this manner, moisture content is calculated. Limitations to the water balloon method
are, again, the length of time needed to get results and also the accuracy depends on
the ability of the balloon to conform to any irregularities along the sides of the hole
2.1.4 Compaction Quality Control in the Field
Current compaction quality control (QC/QA) in the field is fully based on
laboratory compaction testing of soil (Standard Proctor Test). The procedures of the
Proctor Test have been adopted and further standardized by the American Association
of State Highway and Transportation Officials (AASHTO). The Standard AASHTO
procedure (T-99) uses a 5.5 lb. (2.5 kg) hammer dropped freely from a height of 12
inches (3054 mm). Again, the soil is compacted in three layers by 25 hammer blows in a
4 inch (102 mm) diameter mold. This test imparts a total of 12,400 ft. lbs. of compactive
effort to the soil sample.
Modified compaction test is also introduced by AASHTO in connection with
structures require heavier bearing strength to support extremely heavy loads or to limit
14
settlement. According to the Modified AASHTO procedure (T-180), a 10 lb. (4.5 kg)
hammer is dropped from a height of 18 inches (457 mm). The soil sample is compacted
in five layers with 25 blows per layer. The compaction energy is 4.5 times larger than
the Standard AASHTO test that produces 56,200 ft-lbs. of effort.
Laboratory test determines the moisture content at which maximum density can
be attained (figure 2.1). It is realized that this density cannot readily be achieved in the
field by conventional compaction equipment. Therefore, field target densities are
specified as a certain percent of the maximum laboratory dry density. Generally, a
range of 90%-95% of maximum dry density of Standard AASHTO is considered during
field compaction. Likewise, the moisture content must be within a range of the
laboratory determined optimum moisture content.
2.2 Intelligent Compaction Control
FHWA (4) defined the intelligent Compaction (IC) technology as “Vibratory rollers
that are equipped with a measurement/control system that can automatically control
compaction parameters in response to materials stiffness measured during the
compaction process. The roller is also equipped with a documentation system that
allows continuous recordation, through an accurate positioning system, of roller location
and corresponding density-related output, such as number of roller passes and roller-
generated materials stiffness measurements.” The Intelligent Compaction (IC) is made
possible because of the ability of a vibratory roller to first sense the material response of
soil under loading. Then process this information and compare it to the input
requirements. After that, it “decides” how to adjust compaction parameters to most
efficiently compact the material (4, 13). Since none of these features are available on
15
conventional vibratory rollers, IC represents a major innovation in soil compaction
technology (4).
2.2.1 Historical Background
Intelligent compaction is a concept during the last three decades. The concept is
continued to be advanced by BOMAG in Germany, AMMANN in Switzerland and
GEODYNAMIK in Sweden. In 1982, the first measurement system for soil compaction
was introduced by BOMAG. In 1989, the first documentation system for soil compactors
was presented. The German Ministry of Highways Construction gave its first
recommendations on SCCC (Soil Continuous Compaction Control, a pioneer of
Intelligent Compaction) in 1993. In 1994, the same highway agency introduced
specification on SCCC. A volumetric roller was introduced for asphalt in 1996 and a
variocontrol roller for soil in 1998. In the late 1990s, both AMMANN and BOMAG
conducted studies to validate the relationship of the roller measured stiffness of soil and
the material properties related to in-place density levels. The study was successful in
identifying the proportional relationship between the plate load test data and roller-
measured soil stiffness for granular soil when the compacted soil mass reached near to
its optimum density level. Then finally, Ammann Compaction Expert (ACEE) introduced
the modulus Evib in 2000 and correlated its value with plate lad tests in 2001 (4).
Intelligent Compaction is more popular in Europe compared to US because
different types of contracts are used for the construction projects (4). European methods
of contract procurements and administration were very similar to those in the United
States until the late 1980s. Public transportation agencies retained tight control over the
design and construction of the highway systems. In the late 1980s, European agencies
16
began to make significant changes to contract administration technique. During a
European scan tour, FHWA identified certain technologies that could be implemented in
US. High quality embankment contraction was one of them and hence intelligent
compaction idea was introduced by FHWA for field testing in the US environment. As of
now, the technology has been (or being) field tested in a number of states (Figure 2.2).
A survey of US roller manufacturers has concluded that Ammann-America,
Bomag-America, Dynapac US and Caterpillar are the four major compaction equipment
manufacturers. They are actively developing the technologies that have intelligent
related capabilities at this time (4). Some of these companies have already started to
market their technologies in the United State.
2.2.2 Background Mechanics of Intelligent Compaction
It is important to understand the term “material stiffness” as it is the conceptual
basis for intelligent compaction. The stiffness is calculated continuously, around 30 to
60 times per second, as a function of the acceleration or force of the roller drum and the
FIGURE 2.2: Application of Intelligent Compaction in US (7)
17
displacement of the material that is being compacted. In general, stiffness is totally a
function of internal friction or aggregate interlocking of the pavement materials. In this
situation, the underlying material is considered as a rigid or semi-rigid solid foundation.
CONCEPT OF STIFFNESS AND MODULUS
The fundamental definition of stiffness is the ratio of the force applied to the
material by a loaded area and the resulting displacement experienced by the loaded
area. Stiffness (k) is expressed in units of force per unit length (kN/m). Soil modulus is
defined as the applied force per unit area of the loaded plate since it is not the slope of
the stress-strain curves of the material under pressure (6). Its unit is expressed as
(kN/m2). The loaded plate area could be square, circular or in the shape of a ring. For
the case of a circular plate having a diameter of D, the soil modulus can be expressed
as:
E ƒ kD
Equation 2.1
This relation shows that modulus is a material property while stiffness is not a soil
property and depends on the size of the loaded area. Therefore, for an elastic material,
stiffness measurement of a particular material will vary from one test to another. But,
modulus of an elastic material is more or less the same for all tests performed and is the
best way to evaluate the material behavior.
OBTAINING THE STIFFNESS kB USING IC ROLLER
A dynamic soil compactor produces nonlinear oscillations during compaction.
The characteristics properties of a nonlinear oscillation can be described by an
analytical procedure (14). Figures 2.3 and 2.4 describe a theoretical, lumped parameter
18
model of the interaction between a vibratory roller and the underneath material. The
soil-drum-interaction force (FB) is defined as follows:
..
2dB d u u f dF m x m r cos t m m g Equation 2.2
where,
md = mass of the drum (kg);
xd = vertical displacement of the drum (m);
..
dx = acceleration of the drum (m/s2);
mf = mass of the frame (kg);
mu = unbalanced mass (kg);
ru = radial distance at which mu is attached (m);
muru = static moment of the rotating shaft (kg.m); and
Ω =2 f ;
f = frequency of the rotating shaft (Hz)
t = time elapsed, (sec)
g = acceleration due to gravity (m/sec2)
FIGURE 2.3: Theoretical, Lumped parameter model of interaction between
a vibratory roller and underneath material, (15)
19
In equation 2.2, the downward force is always considered as positive and the
inertia force, ..
d dm x is should be directed negatively. If the subsoil is considered as spring
and dashpot system, the soil-drum interaction force can also be expressed as:
.
dB B d BF k x d x Equation 2.3
Where,
kB= Stiffness of the soil, (kN/m);
dB = Damping coefficient, (kN.s/m) (a damping ratio of 0.2 is assumed);
and
.
dx = Velocity of the drum, (m/s).
The acceleration of the roller drum and the phase angle between excitation and
oscillation can also be measured. Measuring oscillation will help to calculate FB by using
equation (2.2). As all the quantities are known on the right hand side of the equation,
the stiffness value can be obtained by combining equation (2.2) and (2.3).
FIGURE 2.4: Loading loop due to applied force on the roller drum, (7)
20
Alternatively, the force settlement curve can be plotted, as shown in Figure 2.5,
and the slope of the curve on the loading portion can be calculated as the dynamic
stiffness of the material being compacted.
OBTAINING THE MODULUS, E
As mentioned earlier, the soil stiffness is a dependent parameter and modulus E
is an independent parameter. Thus it is necessary to obtain soil modulus E from the soil
stiffness value. This problem was solved by Hertz in 1895 and further developed by
Lundberg in 1939. Hertz and Lundberg postulated the relationship between the load on
the roller and the imprint area created by the roller on an elastic half space. Thus this
solution was used to develop the relationship between the stiffness kB and soil modulus,
E.
FIGURE 2.5: Soil reactions vs. roller amplitudes (after 7)
21
The relation between the stiffness and the modulus can be expressed as:
VIBB
32 VIB
2
f d
E Lk
L E12 1 2.14 ln
2 1 16 m m R g
Equation 2.4
Where, L is the drum width, is Poisson‟s ratio, ln is natural logarithm, mf and
md are the masses contributed by the frame and the drum of the roller, R is the radius of
the drum, and g is the acceleration due to gravity. Knowing kB, Equation (2.4) gives EVIB.
The relationship between the stiffness kB and soil modulus, E was also
established on an experimental basis. The experiments were performed by testing the
IC roller and the plate load test in parallel. Figure 2.7 shows the relationship obtained
between the roller stiffness kB and soil moduli ME1 and ME2 form the first load and the
reload of the plate load test.
FIGURE 2.6: Hertz (1895), and Lundberg (1939) solution for modulus (7)
22
2.2.3 Compaction Measurement and Documentation Systems of IC Roller
The intelligent compaction roller has a powerful range of compaction of all types
of material used in the earth and rock works because of the adjustable amplitude. The
instant and continuous adjustment of amplitude and compaction energy produces
maximum compaction per pass and also avoids loosening of the surface especially on
gravel, sand and anti-frost layers. Currently intelligent compaction rollers are marketed
by BOMAG, AMMANN and Geodynamik. This section will briefly describe the working
principles and compaction measurement and documentation system of the BOMAG IC
roller used in this study.
2.2.3.1 Measurement System
The BOMAG measurement system EVIB meter (BEM) and the Terrameter BTM
plus/BTM prof are used as an integrated system for continuous assessment of
compaction. BOMAG compaction measurement systems are used to support the roller
FIGURE 2.7: Relationship between stiffness and modulus (6)
23
operator to optimize the construction schedule as well as surface area dynamic
compaction control. BOMAG measuring system consists of
Two acceleration transducers
Distance measuring unit
Electronic unit
EVIB analogue display (for BEM) and BOMAG operation panel (BOP) as
operation and control unit (for Terrameter BTM plus/BTM prof).
Printer as standard for the Terrameter BTM plus/BTM prof.
The BOMAG BCM 05 compaction management system is used as a supplement
to the BOMAG compaction measurement systems on single drum rollers (BEM, and the
Terrameter BTM plus/BTM prof). BCM 05 includes tablet PC with touch screen
(Pentium III), USB memory stick, 128 MB suitable for measuring 100,000 m2 and BCM
05 mobile and BCM 05 office software.
FIGURE 2.8: BOMAG measurement and documentation system, (2)
24
2.2.3.2 Measurement Principle
BVC (BOMAG VarioControl) single-drum roller is equipped with the BOMAG
Vario Vibration system. This system generates the linear vibration of the roller drum.
The direction of vibration is adjusted progressively between vertical and horizontal
direction. Adjustment of the direction of vibration significantly influences the transmitted
compaction energy. VARIOCONTROL has the capacity to use much larger amplitudes
than the conventional vibratory roller. Figure 2.9 shows the automatic adjustment of the
compaction energy in response to the soil conditions and the display of the compaction
results and documentations.
FIGURE 2.9: BOMAG Intelligent Compaction System, (2)
25
BOMAG compaction measurement systems use the reciprocal effect of the
acceleration of the vibrating drum and the dynamic stiffness of the soil that generally
increases with compaction. The energy required for the compaction of soil is defined by
the soil itself, its layer thickness, the sub-base and the degree of compaction as well as
the resulting soil stiffness (2). With increasing stiffness of the soil, the contact force also
increases. Two accelerator sensors record and evaluate the stiffness developed and
continuously measure the dynamic behavior of the roller drum. This information is used
for the automatic adjustment of the vibration direction of the roller drum.
The measuring systems record the acceleration and determine the contact force
acting between the soil and the drum as well as the vibration amplitude of the roller
body. During application of the contact force, a loading and unloading curve is
developed for each revolution and the enveloped area of the loading-unloading curves
represents the compaction energy released during compaction. The EVIB value is
calculated from the analysis of the loading curve.
The different position of the acceleration transducer is set to maintain the degree
of compaction in the field. The vertical direction gives the maximum energy available for
the roller and used in higher depth of compaction. The angled position results in
moderate compaction output while the horizontal position of the transducer gives the
minimum lever of compaction. The minimum energy available from the horizontal
direction improves compaction close to the surface and reduces loosening (Figure
2.10).
26
FIGURE 2.10: Direction of applied vibration to optimize compaction, (7)
27
CHAPTER 3 - FIELD TESTING AND DATA COLLECTION
3.1 General
FHWA acknowledged that compaction is an important factor for all pavement
materials. Hence, emphasis should be placed on the proper densification of all
pavement layers including subgrade soils, granular base and asphalt pavements. In
addition, an understanding of the compaction model shows that the proper compaction
of each layer depends on the proper densification of the underlying layer. If one or all of
the pavement foundation layers lack sufficient stiffness, it will be more difficult or
sometimes totally impossible to properly compact the layer on top of it.
In recent years, it has been proven that intelligent systems are comprehensive
products where different technologies are integrated into a single product to work
together. Intelligent products can make a conventional product perform its function more
effectively and also can add capabilities to a product through sensing and mechanical
and/or software improvements.
3.2 Description of Test Section
Two pavement reconstruction projects, one on US-56 and one on I-70, were
selected for this study. The US-56 project had two 328 ft (100 m) sections. The first
section was a “proof” section that had already been compacted by a conventional roller.
IC roller stiffness measurements were made after one pass of the roller. The other
section was a “growth” section that was compacted by the IC roller using multiple
passes and densities were “built” up. The I-70 section had only one 328 ft (100 m)
“growth” section. The stiffness on both “growth” sections were measured at each pass
and compared with that obtained for the previous pass. A decrease in stiffness between
28
two consecutive passes was meant to be the end of compaction. The target density and
corresponding moisture content of the “growth” test sections were those required by the
project special provisions. No proof rolling was performed on any of these test sections.
3.2.1 Laboratory Testing
Soil samples were collected at 16.5 ft (5 m) intervals on all sections. The soil
samples were tested for gradation, Atterberg limits (unsuccessfully) and moisture-
density relationships in the laboratory. Soil samples were also collected for in-situ
moisture content determination by the gravimetric method in the laboratory. Typical
results have been tabulated in Table 3.1. Figure 3.1 shows the moisture-density curves
on the three test sections obtained from the standard Proctor tests. Both projects have
sandy (SP or A-3) soils. Other relevant soil characteristics are shown in Table 3.1.
Project US-56 Proof US-56 Growth I-70 Growth
% Passing # 200 Sieve (%) 4.2 6.8 8.4
Coefficient of Uniformity, Cu 2.1 4.7 4.2
Coefficient of Curvature, Cc 1.25 0.67 1.6
Soil Classification (AASHTO) A-3 A-3 A-3
Plasticity NP NP NP
Avg. In-situ Moisture content (%) 6.4 5.6 8.98
Optimum Moisture Content (OMC), (%) 10.4 10.2 11.7
Target moisture content (%) 5.4-15.4 5.2-15.2 6.7-11.7
Maximum Dry Density, MDD, (kg/m3) 1916.0 2012.0 1838.8
Target dry density, (kg/m3) 1820 1911 1747
Poisson‟s ratio 0.35 0.35 0.35
TABLE 3.1: Test Section Soil Characteristics
29
3.3 IC Roller Description
The IC roller used in this study was a Bomag VARIOCONTROL (BVC) single
drum vibratory roller (Model BW 213 DH-4 BVC) as shown in Figure 3.2. The roller had
a gross weight of 31,976 lbs (14,505 kg) and the weight on the single axle, steel drum
was 20,283 lbs (9,200 kg). The working width was 83.9 inch (2130 mm). The speed of
the roller was 0 to 8.1 mph (0-13 kmph). The vibration frequency was 1,680 vpm (28
Hz) and the amplitude was 0 to 0.094 inch (0-2.4 mm). The centrifugal force produced
was 83,215 lbs (365 kN).
FIGURE 3.1: Moisture-Density curves from Standard Proctor Test
US-56 Proof Section
1740
1760
1780
1800
1820
1840
1860
1880
1900
1920
1940
2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0
Water Content, w (%)
Dry
Un
it W
eig
ht,
γd
, k
g/m
3
US-56 Growth Section
1820
1840
1860
1880
1900
1920
1940
1960
1980
2000
2020
6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0
Water Content, w (%)
Dry
Un
it W
eig
ht,
γd
, k
g/m
3
I-70 Growth Section
1720
1740
1760
1780
1800
1820
1840
1860
4.0 6.0 8.0 10.0 12.0 14.0 16.0
Water Content (%)
Dry
Un
it W
eig
ht,
γd
, k
g/m
3
30
In this study, EVIB data was continuously collected over the test sections. On the
“proof‟ section on US-56, BVC stiffness data was collected after one pass of the roller.
For the “growth” sections, EVIB data was collected for each pass of the roller.
3.4 In-situ Testing and Data Collection
In-situ testing was performed for the quality control of compaction. It involved
some spot tests such as, Geogage for stiffness measurement, Falling Weight
Deflectometer (FWD) and Light Falling Weight Deflectometer (LFWD) for modulus
testing, the Nuclear Gage for density and in-situ moisture testing, and Dynamic Cone
Penetrometer (DCP) for measuring the California Bearing Ratio (CBR). These methods
are routinely used in compaction control and/or measuring soil stiffness characteristics.
FIGURE 3.2: BVC roller compaction on a test section on US-56 near
Hugoton, Kansas
31
3.4.1 Density and Moisture Measurements
The in-situ density measurements were done by a nuclear gage (Figure 3.3).
Measurements were taken at 16.5 ft (5 m) intervals on all sections. In-situ moisture was
also measured on those test sections locations using the nuclear gage. As mentioned
earlier, soil samples were also collected for in-situ moisture content determination by
the gravimetric method in the laboratory.
3.4.2 Stiffness, Deflection, and Cone Penetration Tests
The Geogage, Falling Weight Deflectometer (FWD), Light Falling Weight
Deflectometer (LFWD) and Dynamic Cone Penetrometer (DCP) measurements were
made at 16.5 ft (5 m) intervals. For the US-56 “proof” section, tests were done after one
pass of the BVC roller. For the “growth” sections, tests were done after the final pass of
the BVC roller.
FIGURE 3.3: In-situ moisture and density measurement using nuclear
gage on I-70
32
3.4.2.1 Geogage
The soil stiffness gage or Geogage is a portable, nondestructive, and non-
nuclear device that was used to measure the soil stiffness. It was 0.8 ft (250 mm) in
height, resting on a 0.9 ft (280 mm) diameter base, and weighed about 22 lbs (10 kg).
The base was a rigid ring-shaped foot on the soil surface. The applied force and the
displacement-time history were measured by two velocity sensors. According to the
manufacturer, the geogage vibrates and
produces small changes in vertical force
with displacement within the 6,000 to
12,000 vpm (100 to 200 Hz) frequencies.
The in-situ soil stiffness was measured at
each frequency and then finally, the
average value was recorded. It
measured stiffness up to 0.7 to 1 ft (220
to 310 mm) of depth from the contact
surface. The geogage stiffness can also
be used to determine the Young‟s
modulus (16). Figure 3.4 shows the in-
situ stiffness measurement on US-56
“growth” section after the final pass of
the IC roller.
FIGURE 3.4: Stiffness measurements by
Geogage, on US-56 section
33
3.4.2.2 Falling Weight Deflectometer (FWD)
The Dynatest model 8000 FWD system was used in this study to obtain the in-
situ deflection data. The impulse force was created by dropping a target mass of
2500±200 lbs (1134±91 kg) from a certain height. This load level was recommended by
George (17) for bare subgrade testing using an FWD. The load was transmitted to the
subgrade through a load plate with a diameter of 18 inch (457 mm) to provide a load
pulse in the form of a half sine wave with a duration of 25 to 30 ms. The load magnitude
was measured by a load cell. Figure 3.5 shows the deflection measurements made by
FWD device on the US-56 “proof “section.
Deflections were measured using seven sensors mounted on a bar that was
lowered on to the pavement surface automatically with the loading plate. One of the
sensors was located at the center of the plate while the other six were located at a
radial distance of 12 inch (305 mm) center to center. The measured deflections at
different stations were used to back-calculate the modulus of the subgrade soil. There
FIGURE 3.5: Deflection measurements on US-56 section by FWD
34
are some general techniques that match the measured deflections with those calculated
from theory. Example computer programs that make use of this technique include
EVERCLAC 5.0, MODCOMP and MODULUS 6.0. In this study, EVERCALC 5.0 was
used to backcalculate the subgrade modulus from the in-situ deflection data. In most
cases, full deflection basin was used. In several instances, the seventh sensor data was
ignored since the deflection recorded by this sensor was more than the sixth sensor.
Figure 3.6 shows the normalized deflection basin plotted during backcalculation.
In a few cases, the whole basin was discarded when the root-mean square
(RMS) error in backcalculation was deemed too high. The first sensor deflection was
also used to backcalculate the subgrade soil elastic modulus from the Boussinesq
equation shown later.
FIGURE 3.6: Normalized deflection basins on US-56 section (station
15+00) by EVERCALC 5.0
35
3.4.2.3 Light Falling Weight Deflectometer (LFWD)
Recently hand-held, falling weight deflectometer devices have become available
for surface deflection measurements. The Prima 100 LFWD device was used to
evaluate the in-situ soil modulus in this study. Figure 3.7 shows the LFWD testing on
the I-70 test section to measure in-situ deflection.
The device consisted of four major parts: sensor body, loading plate, buffer
system, and the sliding weight. The
sensor body enclosed both the load cell
and the geophone. The loading plate,
buffer system and the sliding weight
were attached to the sensor body. A 12
inch (300 mm) diameter steel loading
plate, which also doubled as the sensor
body, was used in this study. The
LFWD device measured both force and
deflection. The elastic modulus of the
subgrade soil was calculated from the
surface deflection using the following
Boussinesq equation: FIGURE 3.7: Deflection measurements by
LFWD on I-70 section
36
2
o
LFWD
o
k 1 aE (0)
d Equation 3.1
Where,
LFWDE = LFWD modulus (MPa);
k = 2
and 2 for rigid and flexible plate, respectively;
d o = deflection at center (μm);
o = Applied stress (kPa);
= Poisson‟s Ratio; and
a = plate diameter (mm).
In this study, a rigid plate was assumed during modulus calculation from the
Boussinesq analysis.
3.4.2.4 Dynamic Cone Penetrometer (DCP)
Dynamic Cone Penetrometer was initially developed in South Africa as an in-situ
pavement evaluation technique for continuous measurement with the depth of
pavement layers and subgrade soil parameters (18). Since then this device has been
used extensively in South Africa, United Kingdom, US, Australia and many other
countries because it is simple, economical, and less time consuming than most other
available methods. The DCP used in this study was provided by Managing Technology,
Inc., Overland Park, Kansas to KDOT in the early nineties. The KDOT DCP consisted of
a slender steel rod with a cone tip at the end (Figure 3.8). The cone tip was made of
hardened steel and was angled at 300 with a diameter at its head of 0.8 inch (20 mm).
The hammer which slided down the steel rod had a height of 23 inch (575 mm) and
weighed 18 lbs (8 kg). The unit had two aluminum blocks and a reference beam that
37
aided in measuring the penetration depth during testing. For subgrade evaluation in this
study, the DCP was penetrated down from the top of the compacted subgrade. During
testing, the number of blows vs. depth was recorded. The "DCP value" was defined as
the slope of the blow vs. depth curve (in mm per blow) at a given linear depth segment.
Correlation between DCP and CBR: The California Bearing Ratio (CBR) test
measures the static penetration resistance of a soil as a function of penetration of a
cylinder prior to reaching the ultimate shearing value of the soil. The CBR is defined as
a percentage determined by the ratio of the resistance in psi at 0.1 inch (2.5 mm)
penetration of the soil under test to the resistance of a standard, well graded, crushed
stone at the same penetration .1 inch (2.5 mm), and then multiplied by 100. This
standard penetration stress is usually taken to be 1,000 psi (6,895 kPa).
FIGURE 3.8: DCP testing on I-70 test section in Kansas
38
In order to assess the structural properties of the pavement subgrade, the DCP
values are usually correlated with the CBR of the pavement subgrade soil (19):
1.5
logCBR 220 0.71 logDCP Equation 3.2
(R2 = 0.95, N = 74)
where the DCP is in mm per blow.
This relationship was verified for a wide range of pavement and subgrade
materials (18). Additional laboratory and correlation work conducted at the University of
Kansas (19) and by the U.S. Army Corps of Engineers (Waterways Experiment Station)
(20) generally supported the relationship described in Equation (3.2), but indicated
considerable data scatter. It was recommended the DCP limit of 1.0 inch/blow (25
mm/blow) be correlated to a CBR of 8 for the silty/clay soils although the actual DCP
value of the soil was way above this limit (14). After calculating the CRB values, the
modulus of soil was calculated using CBR-modulus relationship proposed by M-EPDG
(21). The proposed relationship is as follows:
0.64
rM 2555 CBR Equation 3.3
Where,
rM =Resilient modulus of soil in psi
CBR = California Bearing Ratio
39
CHAPTER 4 - RESULTS AND DISCUSSION
4.1 General
As mentioned before, the in-situ stiffness of embankment soil was measured by
Bomag IC roller. IC roller reports continuous measurement of soil stiffness along the
entire test section. Stations at 16.4 ft (5 m) interval were marked along the test sections.
Thus each 328 ft (100-m) test section had 21 test locations for discrete measurements
by Geogage, FWD, LFWD, DCP and the nuclear gage. The Geogage measured soil
stiffness values. Soil density and moisture measurements were obtained with the
nuclear gage and speedy moisture tester respectively. The modulus values from the in-
situ deflections measured by FWD were backcalculated using EVERCALC 5.0
backcalculation software. The modulus values were also backcalculated from the
Boussinesq equation using the first sensor deflection data. This equation was again
used to backcalculate the modulus from the LFWD deflection data. DCP values were
used to calculate CBR values from the DCP-CBR correlation developed by the Corps of
Engineers. Then the modulus value was calculated using the CBR-Modulus relationship
in M-EPDG.
The variation of soil stiffness measured by the IC roller was examined along the
test section. Stiffness variation with in-situ moisture content and percent compaction
was also investigated. A statistical analysis was performed to obtain the point statistics
of the stiffness/modulus value obtained from different testing methods. A correlation
matrix among these stiffness/modulus values was also developed using the statistical
software SAS. The following sections present the analysis and discussion of the results
obtained from different tests.
40
4.2 BVC Stiffness on US-56 “Proof”, “Growth” and I-70 “Growth” Sections
BVC stiffness (EVIB) measurements were made after a single pass on a 328 ft
(100 m) “proof” section on US-56 that had been compacted by a conventional vibratory
roller. Other BVC stiffness measurements were taken on 328-ft (100 m) “growth” section
on both US-56 and I-70 that had been compacted by the multiple passes of an IC roller.
The results are shown in Figure 4.1.
FIGURE 4.1: Stiffness developed and measured by IC roller on I-70 and
US-56
BVC Stiffness
US-56 "Growth" Section
0
10
20
30
40
50
60
+00 2+00 4+00 6+00 8+00 10+00 12+00 14+00 16+00 18+00 20+00
Station
Sti
ffn
ess
(MN
/m2 )
Initial Intermediate Final
Bomag Stiffness
I-70 "growth" Section
0
10
20
30
40
50
60
70
80
+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Station
Stiff
ness
(MN/
m2 )
First Pass Second Pass Third Pass Fourth Pass
41
The results in Figure 4.1 show that in general, density increases with multiple
passes of the IC roller, but at a few locations density decreases. This phenomenon may
happen due to presence of excess moisture at these locations.
On the US-56 “proof” section, mean IC stiffness (EVIB) was 9 ksi (62 MPa) and it
varied from 4 ksi (30 MPa) to 18 ksi (120 MPa). US-56 “growth” section showed a mean
stiffness value of 5 ksi (36 MPa) with the range being from 2.9 ksi (20 MPa) to 7.25 ksi
(50 MPa). On I-70, the mean EVIB was 6 ksi (40 MPa), and it varied from 4 to 9 ksi (30 to
60 MPa). As shown in Figure 4.2 indicates, there were a number of soft spots along
these small stretches of the embankment. Due to continuous measurement of the soil
stiffness, it was possible for the IC roller to identify these locations.
(a)
0
20
40
60
80
100
120
140
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Stations
Sti
ffn
ess, (M
N/m
2)
IC roller stiffness Mean stiffness
42
(b)
(c)
FIGURE 4.2: Spatial variation of IC roller stiffness on (a) US-56 “proof”, (b)
US-56 “growth” and (c) I-70 “growth”
0
10
20
30
40
50
60
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Stations
Sti
ffn
es
s, (M
N/m
2)
IC roller stiffness Mean stiffness
0
10
20
30
40
50
60
70
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Stations
Sti
ffn
ess, (M
N/m
2)
IC roller stiffness Mean stiffness
43
4.3 Variation of IC Roller Stiffness with Moisture Content
Subgrade soil is susceptible to moisture variation after construction of the
highway pavement. Briaud and Seo (6) have indicated that the soil at lower moisture
content will have higher modulus values. This effect is more pronounced for the clay-
type soil. On the other hand, density gives the compactness of the soil particles and
determines how they are arranged in a given volume. But, unfortunately there is no
correlation between the soil density and modulus, and the same density can be
obtained for at least two different moisture contents (on either side of the standard
Proctor compaction curve). That is why it is not possible to control soil compaction on
the basis of the dry density alone. Figure 4.3 shows the relationship between the IC
stiffness and the in-situ moisture content for the test sections on US-56 and I-70. The
trends in these graphs clearly indicate that the IC roller stiffness was somewhat
sensitive to the field moisture content. Higher moisture content resulted in lower
vibration modulus.
44
(a)
(b)
0
30
60
90
120
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Station
Sti
ffn
es
s, (M
N/m
2)
0
2
4
6
8
10
12
w, (%
)
IC Stiffness w
0
10
20
30
40
50
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Station
Sti
ffn
es
s, (M
N/m
2)
0
1
23
4
5
67
8
9
w, (%
)
IC Stiffness w
45
(c)
In order to study the variation of in-situ stiffness developed by IC roller with field
moisture content further, the moisture differential curve was also plotted. The moisture
differential (Δw) was calculated by subtracting the field moisture content from the
optimum moisture content (OMC) obtained in the laboratory from the standard Proctor
test. Figure 4.4 shows the relationship between the IC roller stiffness and the deviation
in in-situ moisture content (from OMC) for the test sections. The positive difference
indicates the in-situ moisture content was below the optimum moisture content. The
trends in these graphs clearly indicate that the IC roller stiffness, in general, is sensitive
to the field moisture content. At moisture contents near the optimum, IC roller vibration
modulus was generally higher.
FIGURE 4.3: Variation of IC roller stiffness with in-situ moisture content on
(a) US-56 “proof”, (b) US-56 “growth” and (c) I-70 “growth” section.
0
10
20
30
40
50
60
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Station
Sti
ffn
es
s, (M
N/m
2)
0
2
4
6
8
10
12
w, (%
)
BVC Stiffness w
46
(a)
(b)
0
30
60
90
120
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Station
Sti
ffn
es
s, (M
N/m
2)
-1
0
1
2
3
4
5
6
7
Δw
, (%
)
IC Stiffness Δw
0
10
20
30
40
50
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Station
Sti
ffn
es
s, (M
N/m
2)
0
1
2
3
4
5
6
7
Δw
, (%
)
IC Stiffness Δw
47
(c)
FIGURE 4.4: Variation of IC roller stiffness with moisture differential on (a)
US-56 “proof”, (b) US-56 “growth” and (c) I-70 “growth” section.
0
10
20
30
40
50
60
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Station
Sti
ffn
es
s, (M
N/m
2)
0
1
2
3
4
5
6
Δw
, (%
)
BVC Stiffness Δw
48
4.4 Variation of IC Stiffness with Compaction Level
The relationship between the in-situ dry density (obtained after compaction and
expressed in terms of the maximum dry density from the standard Proctor test), and IC
roller stiffness was also examined. This was done due to the fact that the current
practice of embankment compaction control in Kansas is based on the percent
compaction obtained from the maximum dry density (MDD) value in the standard
Proctor test. The in-situ dry density is expressed in terms of percent compaction using
the following relations:
wd in situ
1 w Equation 4.1
Here,
d in situ =In-situ dry density
w = In-situ wet density
w =In-situ moisture content (%)
Then, the % compaction is expressed as:
d in situ
d SPT
%compaction Equation 4.2
Where, d SPT = Laboratory dry density from Standard Proctor Test
49
(a)
(b)
0
20
40
60
80
100
120
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Section
IC R
olle
r S
tiff
ne
ss
, (M
N/m
2)
75
80
85
90
95
100
105
% C
om
pa
cti
on
IC Roller Stiffness % Compaction
0
10
20
30
40
50
60
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Section
IC R
olle
r S
tiff
ne
ss
, (M
N/m
2)
75
80
85
90
95
100
105
% C
om
pa
cti
on
IC Roller Stiffness % Compaction
50
(c)
Figure 4.5 shows the relationship between the IC roller stiffness and percent
compaction obtained from the two sections on US-56 and I-70. The US-56 results show
that lower IC roller stiffness was obtained for both very high and very low percent
compaction. Although the I-70 test section showed higher levels of percent compaction,
the trend was almost similar to that observed on the US-56 test section. This finding is
very significant since it indicates the need for developing the “target” stiffness for the IC
rollers for a specific type of soil. When the preselected stiffness value is entered prior to
the compaction, the IC roller automatically controls the compaction process until the
target stiffness value is obtained. At that point, the roller reduces or eliminates the
compactive effort on subsequent passes. If these target values are low, the resulting
FIGURE 4.5: Relationship between IC roller stiffness and percent
compaction for (a) US-56 “proof”, (b) US-56 “growth” and (c) I-70 “growth”
sections
0
10
20
30
40
50
60
70
0+00
2+00
4+00
6+00
8+00
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Section
IC R
oll
er
Sti
ffn
es
s,
(MN
/m2
)
92
94
96
98
100
102
104
106
108
% C
om
pa
cti
on
IC Roller Stiffness % Compaction
51
density will be too low where as high target values will result in overcompaction. The
results of this study tend to prove this point.
4.5 Variation of IC Roller, Geogage, LFWD, and FWD Stiffness
FHWA (4) reported that some European and Asian countries now use
compaction specifications that contain modulus-based compaction control. In those
specifications, minimum target modulus must be obtained in addition to the target
moisture content and percentage of the Proctor density. The countries that have
switched to modulus-based compaction control have typically compared their roller
modulus with the field plate loading test modulus. The specifications vary from one
country to another but they only vary in their minimum target modulus values depending
on traffic, material type, and subgrade soil classification. For example, typical values of
roller stiffness in one European specification have been reported to be 6.5 ksi (45 MPa)
and 18 ksi (120 MPa) for low traffic roads and freeways, respectively (6). In the United
States, the plate modulus test (or plate loading test) is not used by all states as a
standard acceptance tool. However, Geogage, Dynamic Cone Penetrometer (DCP),
light falling weight deflectometer (LFWD) and falling weight deflectometer (FWD) have
become popular tools for subgrade evaluation. Table 2 shows the test conditions for
various tests done on the projects in this study. The table also shows the mean stiffness
and moduli obtained from the IC roller and other tests. It is to be noted that all tests
were done at different vertical pressure. It is well known that the subgrade soil is stress
dependent –the soil modulus varies with the deviator stress and to some extent,
confining pressure. For sandy soils, the modulus increases with increasing deviator
stress. Thus the layer modulus changes with depth. This is potentially a source of
52
problem when comparing stiffness and modulus results from different tests. In this
study, efforts were made to keep the applied vertical stresses of FWD and LFWD
similar.
Test
Section Test
Applied
Vertical
stress, ζd
(kPa)
Field
Moisture
Content
w (%)
OMC
w (%)
Average
Stiffness/Modulus
(MPa)
US-56
“proof”
IC Roller 92.6-277.8
6.2 – 9.9 10.4
61.7
Geogage 25.0 7.96
LFWD 130.9 – 162.5 64.1*
FWD 67.6 – 116.5 73.3*
US-56
“growth”
IC Roller 92.6-277.8
4.05 – 7.9 10.19
36.4
Geogage 25.0 11.4
FWD 130.9 – 162.5 72.4*
I-70
“growth”
IC Roller 92.6-277.8
6.2 – 9.9 11.7
40.5
Geogage 25.0 4.9
LFWD 118.3 – 136.4 26.4*
FWD 52.4 – 102.73 29.8*
* from Boussinesq equation
TABLE 4.1: Applied Vertical Stresses and Moisture Content During IC
Compaction and Testing
53
(a)
(b)
0
50
100
150
200
250
0+00
02+0
0
04+0
0
06+0
0
08+0
0
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Section
IC, L
-FW
D, F
WD
Sti
ffn
es
s,
(MN
/m2
)
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
Ge
og
au
ge
Sti
ffn
es
s, (M
N/m
2)
IC L-FWD FWD Geogauge
0
20
40
60
80
100
120
140
160
180
200
0+00
02+0
0
04+0
0
06+0
0
08+0
0
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test Section
IC, F
WD
Sti
ffn
es
s, (M
N/m
2)
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
16.00
18.00
20.00
Ge
og
au
ge
Sti
ffn
es
s, (M
N/m
2)
IC FWD Geogage
54
(c)
Figure 4.6 illustrates the variation of various stiffness and moduli obtained at
different test locations on the US-56 “proof” and “growth” and I-70 “growth” sections.
The US-56 “growth” section did not have any LFWD data. The highest variation was
obtained for the IC roller stiffness and the LFWD-derived modulus. The trend of data in
Figure 4.6(a) shows a good correlation between the Geogage stiffness and the LFWD
backcalculated modulus on the US-56 “proof” section. On this section, the mean
stiffness and moduli values obtained from the IC roller, LFWD and FWD (Boussinesq‟s
equation) were also somewhat close.
However, the above trend is not evident on the US-56 and I-70 „growth” sections.
As expected, the mean modulus values from LFWD and FWD were similar. However,
as illustrated in Figure 4.6(b) and 4.6(c), no definite correlation is evident among the IC
FIGURE 4.6: Variation of IC roller, FWD, LFWD and Geogage stiffness on (a) US-
56 “proof”, (b) US-56 “growth” and (c) I-70 “growth” sections
0
10
20
30
40
50
60
70
0+00
02+0
0
04+0
0
06+0
0
08+0
0
10+0
0
12+0
0
14+0
0
16+0
0
18+0
0
20+0
0
Test section
IC, L
-FW
D, F
WD
Sti
ffn
es
s,
(MN
/m2
)
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
Ge
og
au
ge
Sti
ffn
es
s, (M
N/m
2)
IC L-FWD FWD Geogauge
55
stiffness, Geogage stiffness, and LFWD and FWD backcalculated moduli. Further
statistical analysis was performed with the stiffness and modulus values calculated from
different test methods to check correlation among the different measures.
4.6 Statistical Analysis of Test Results
A correlation matrix was developed to investigate the correlation among the
stiffness values. Table 4.2 tabulates the statistical summary of the IC roller stiffness and
other stiffness and moduli results obtained in this study. IC roller stiffness and Geogage
stiffness were the outputs of these pieces of equipment. Boussinesq‟s equation was
used to backcalculate the soil elastic modulus from both LFWD and FWD in-situ
deflection data. FWD modulus was further calculated using the EVERCALC 5.0
backcalculation software using the full deflection basin from all seven or six sensors.
The CBR/DCP modulus was calculated from the correlation between CBR and the
resilient modulus.
Statistical summary shows that the stiffness values are highly variable on US-56
“proof” and “growth” sections. The highest coefficient of variation (COV) of around 38%
was obtained for the IC roller and LFWD device on the US-56 proof section. The highest
COV value (67%) was observed for the FWD modulus on the US-56 “growth” section.
However, the above trend was not evident on the I-70 “growth” section as shown in
Table 4.2. The stiffness and modulus values were much less scattered on this section.
As expected, the mean modulus values from LFWD and FWD were similar. The
maximum stiffness value was obtained from the DCP test on both US-56 “proof” and I-
70 “growth” section. The minimum stiffness value was measured by Geogage.
56
Test
Section Parameter
Mean
(MPa)
Std. Dev.
(MPa)
Coeff.
Of Var.
(%)
Range
(MPa) n
US-56
“Proof”
BVC Stiffness 61.7 23.2 37.5 90.0 21
Geogage
Stiffness 7.96 2.2 27.4 10.96 21
LFWD Modulus* 64.1 24.1 37.6 91.3 19
FWD Modulus** 86.5 29.7 34.4 135.9 20
FWD Modulus * 73.3 19.2 26.2 87.8 21
CBR/DCP
Modulus*** 129.6 28.3 21.8 98.8 21
US-56
“growth”
BVC Stiffness 36.43 8.24 22.61 30 21
Geogage
Stiffness 11.40 2.96 25.97 11.36 21
FWD Modulus** 79.73 53.03 66.51 176.85 20
FWD Modulus * 72.41 40.17 55.48 130.05 21
CBR/DCP
Modulus*** 125.69 17.88 14.22 61.65 21
I-70
“Growth”
BVC Stiffness 40.5 7.7 19.1 30.0 21
Geogage
Stiffness 4.91 1.1 23.1 5.3 21
LFWD Modulus* 26.4 7.2 27.4 28.1 21
FWD Modulus** 29.1 6.6 22.7 26.2 16
FWD Modulus * 29.8 6.4 21.3 23.8 21
CBR/DCP
Modulus*** 97.3 8.2 8.4 34.2 21
* from Boussinesq equation; **from backcalculation; *** from correlation
TABLE 4.2: Statistical Summary of Stiffness and Moduli Results
57
4.6.1 Correlation Matrix
A Correlation matrix describes correlation among M numbers of variables. It is a
square symmetrical (M x M) matrix with the (ij)th element equal to the correlation
coefficient ij between the (i)th and the (j)th variable. The diagonal elements (correlations
of variables with themselves) are always equal to 1.00. Correlation coefficients can
range from -1.00 to +1.00. The value of -1.00 represents a perfect negative correlation
while a value of +1.00 represents a perfect positive correlation. A value of 0.00
represents a lack of correlation. A correlation matrix is always a symmetric matrix
The correlation among the soil stiffness obtained from the IC roller and Geogage,
and the subgrade soil moduli obtained from the LFWD, FWD, and DCP (with CBR
correlation) data was also examined statistically. Correlation tables were developed
using the SAS statistical software (22)
IC Geogage LFWD FWD
(Backcalculated)
FWD
(Boussinesq
equation)
CBR/DCP
IC 1.0 -0.35
0.12*
-0.39
0.096*
-0.19
0.43*
-0.21
0.36*
-0.44
0.04*
Geogage 1.0 0.78
<0.0001*
0.24
0.31*
0.83
<0.0001*
0.72
0.0002*
LFWD 1.0 0.11
0.68*
0.63
0.004*
0.86
<0.0001*
FWD
(Backcalculated) 1.0
0.36
0.12*
0.18
0.45*
FWD
(Boussinesq
equation)
1.0 0.47
0.033*
CBR/DCP 1.0
* p-value
TABLE 4.3: Correlations among IC and Geogage Stiffness, LFWD and FWD Moduli, and CBR on
US-56 “Proof” Section
58
IC Geogage FWD
(Backcalculated)
FWD
(Boussinesq
equation)
CBR/DCP
IC 1.0 0.04 0.30 0.47 -0.13
Geogage 1.0 0.36 0.12 0.34
FWD
(Backcalculated) 1.0 0.90 0.58
FWD
(Boussinesq
equation)
1.0 0.42
CBR/DCP 1.0
IC Geogage LFWD FWD
(Backcalculated)
FWD
(Boussinesq
equation)
CBR/DCP
IC 1.0 0.39
0.085*
-0.001
0.99*
-0.03
0.93*
-0.05
0.82*
0.05
0.83*
Geogage 1.0 0.32
0.16*
0.037
0.89*
-0.04
0.86*
0.27
0.24*
LFWD 1.0 0.008
0.98*
-0.08
0.75*
-0.08
0.73*
FWD
(Backcalculated) 1.0
0.97
<0.0001*
-0.44
0.087*
FWD
(Boussinesq
equation)
1.0 -0.26
0.26*
CBR/DCP 1.0
* p-value
TABLE 4.4: Correlations among BVC, Geogage, FWD Moduli, and CBR on US-56
“Growth” Section
TABLE 4.5: Correlations among BVC, Geogage, LFWD and FWD Moduli, and CBR on I-70
“Growth” Section
59
Results in Table 4.3 for the US-56 “proof” section tend to confirm the observation
in Figure 4.6(a). The table shows about 78% dependency of Geogage in-situ stiffness
on the LFWD stiffness at 95% confidence level with a p-value less than 0.0001. The
Geogage stiffness also has a good correlation with the modulus obtained from CBR in
the DCP test. However, results in Table 4.5 for the I-70 section show about 32%
dependency of the Geogage stiffness on the LFWD modulus at 95% confidence level
with a p-value equal to 0.1607. None of the test results seems to have a strong
correlation with the IC stiffness on this section. Further testing may be necessary to
develop such a correlation. It may be mentioned that the IC roller stiffness has been
successfully correlated with the plate loading test results in Europe. It appears that
larger plated helps to test a large soil sample that could be considered somewhat
representative of the soil subgrade.
4.6.2 Reasons for Poor Correlation among Different Measures of Stiffness and
Moduli
The poor correlations among the IC roller-generated soil stiffness and stiffness
and moduli measured or derived from the soil stiffness gage, LFWD, and FWD was
further investigated by the volume of influence space of each measurement technique in
the subgrade. It is to be noted that the differences in stress states of these measures
are shown in Table 4.1 and had been discussed earlier. In this section, based on the
area of the foot print of the soil stiffness gage, LFWD, and FWD, and the approximate
sensing depth for each piece of equipment (23), the volume of influence space in the
subgrade was calculated. Figure 4.7 illustrates the volume of influence space
schematically for each measure.
60
It shows that the soil stiffness gage had the lowest influence space of 0.1 cft
(0.0003 cubic meter) whereas the IC roller influence space volume was 200 times of
that. It is obvious that each measure was “sensing” a separate volume of soil, and the
IC roller was clearly by large had the highest sample size among all measurement
methods. This partly explains why a universal correlation among the outputs of all these
pieces of equipment was not found. However, it is possible that if the influence space
volume of any of the measurement method can be increased, a better correlation with
the IC roller stiffness may be found. FWD seems to be the logical choice since it can
accommodate a larger diameter plate (about 30 inches or 762 mm). In that case, the
volume of influence space of FWD would increase by almost three times.
FIGURE 4.7: Volume of influence space in subgrade soil
61
CHAPTER 5 - MECHANISTIC-EMPIRICAL PAVEMENT
DESIGN (M-E PDG) ANALYSIS
5.1 General
Mechanistic Empirical Pavement Design Guide (M-E PDG) is an advanced
pavement design and analysis methodology that has been developed by NCHRP and is
being implemented nationwide. M-E PDG analysis method requires an input of either
estimated or measured moduli for each of the pavement materials in the roadway cross-
section. This design procedure is integrated into a software program. In this study,
version 1.0 of this software was used.
5.2 M-E PDG Design Approach
The M-E PDG is a mechanistic empirical design procedure based on elastic layer
theory of the pavement analysis. It is a complex pavement analysis tool with many input
factors to characterize the pavement materials, climate, traffic and the construction. The
design approach followed in M-EPDG is briefly illustrated in Figure 5.1. This is an
iterative process and includes the following basic steps:
1. The trial design is developed by the designer using several input values such as
traffic, foundation properties, climate, and material properties.
2. The software estimates the damage and key distresses over the design life.
3. The design is verified against the performance criteria at a preselected level of
reliability. The design may be modified to meet performance and reliability
requirements of a particular project.
62
5.3 Overview of the Design Process of M-EPDG
M-EPDG uses an innovative trial approach for selecting design inputs. Figures
5.2 and 5.3 illustrate the overall design process for flexible and rigid (JPCP) pavement
section. In M-EPDG design process, traffic spectra, materials, and climatic factors are
combined with structural elements to develop a trial design. Structural foundation
includes different layer arrangement. Properties, design and construction features of
HMA (flexible pavement)/ PCC (rigid pavement) and other paving materials are also
needed in trial input section. Pavement responses to the combined effects of dynamic
FIGURE 5.1: Design framework of M-EPDG 2000, (1)
Inputs
Structure Materials Traffic
Climate
Selection of Trial Design
Structural Response (ζ, ε, δ)
Damage Accumulation with Time
Calibrated Damage-Distress
Models
Distress Smoothness
Performance
Verification
Failure Criteria
Design
Re
lia
bili
ty
Design
Requirements
Satisfied?
Final Design
No
Re
vis
ed
Tri
al D
es
ign
63
traffic load and climate are computed using finite element and elastic layer computer
models. Cumulative damage in the pavement over the design life of the structure is
calculated using incremental damage approach. The design life is divided into two-week
time periods for flexible pavements and one-month for rigid pavements. In each time
increment, the daily, seasonal, and long-term changes in material properties, traffic, and
climate are considered. The accumulated damage is then evaluated based on the
failure criteria which were established based on acceptable pavement performance at
the end of the design life. The distress types considered in M-EPDG are rutting, fatigue
cracking, and thermal cracking in asphalt-surfaced pavements. The joint faulting and
transverse cracking in jointed plain concrete pavements, and punchouts in continuously
reinforced concrete pavements are considered as distresses in design guide. In
addition, pavement smoothness is considered through the commonly used International
Roughness Index, or IRI. Expected performance is also evaluated at the given reliability
level. Iteration continues if the design does not meet the established criteria.
64
FIGURE 5.2: M-EPDG design process for flexible pavement, (1)
FIGURE 5.3: M-EPDG design process for JPCP, (1)
Traffic Foundatio
n
Climate Materials
Pro
pert
ies Inputs
Trial Design Strategy
Pavement Analysis
Models
Distress Prediction
Model
Damage
Accum
ulation
Meet
Performanc
e
Criter
ia
Modify
S
tr
a
t
e
g
y
Constructabilit
y Issue
Variable
Alternatives
Life Cycle
Cost
Analys
is Select Strategy
Yes
No
Analysis
Strategy
Selection
Site Factors
Trial Design
Design Criteria
Input Processing
Traffic Projections
EICM Analysis of Climate Data
E to K Conversion
Base Modulus and Bond
Condition
PCC Properties
Bottom-Up Cracking
Prediction
Calculate Stress
Calculate Damage
Predict Cracking
Top-Down Cracking
Prediction
Calculate Stress
Calculate Damage
Predict Cracking
Top-Down Cracking
Prediction
Calculate Stress
Calculate Damage
Predict Cracking
IRI
Initial IRI
Cracking, faulting,
spalling
Site Factors Total Cracking
Requiremen
t
Satisfied?
Design
Revise Trial Design
No
Yes
65
5.4 Test Sections
As mentioned before, two pavement reconstruction projects, one on US-56 and
one on I-70, were selected for this study. The project on US-56 began at the Morton-
Stevens County Line and extended to the east of the west city limit of Hugoton. The
project had two 328 ft (100 m) test sections. The other project, located on Interstate
route 70, began at the Salina-Dickinson County Line and extended to 2.7 km east of RS
187. A test section of 328 ft (100 m) length was selected at this project location.
The proposed design alternate for the reconstruction of US-56 test section is a
full depth asphalt concrete pavement. For the I-70 section, both full depth asphalt
concrete pavement and Portland cement concrete pavement (PCCP) are planned as
two design alternates. The design analysis to get the cross sections was performed by
the 1993 AASHTO Guide for Design of Pavement Structures. The design subgrade
resilient moduli were obtained from the laboratory testing. Figure 5.4 and Table 5.1
show the design cross sections and details of the test sections respectively.
66
Project ID Route Traffic
Direction
Pavement
type
Thickness
(inch)
Mix
Design
Binder
Grade
56-095 K-
6400-01 US-56 EB
Full Depth
AC 13.4
SM-9.5A 70-28*
SM-19A 70-28*
SM-19A 64-22**
70-21 K-
6794-01 I-70 EB
Full Depth
AC 18.1
SM-9.5A 70-28*
SM-19A 70-28*
SM-19A 64-22**
70-21 K-
6794-01 I-70 EB JPCP 12.6 - -
*Binder Grade of Wearing course and Asphalt binder ; ** Binder Grade of Asphalt base
FIGURE 5.4: Full Depth Flexible Pavement (US-56, I-70 Section) and PCCP
(I-70) layered system
TABLE 5.1: Project Details of the Test Sections
Asphalt Surface
Asphalt Binder
Asphalt Base
Compacted Subgrade
Natural Subgrade
PCC
PCTB
Compacted Subgrade
Natural Subgrade
Full depth Asphalt Concrete PCC pavement (PCCP)
67
The JPCP on I-70 has a 15 ft (4.6 m) joint spacing with dowel bars of 1.575
inches (40 mm) diameter. The PCC slab will be constructed on Portland cement
stabilized base and lime treated subgrade. The AASHTO soil classification of the
natural subgrade soil type is primarily Non-Plastic A-3. The top 6 inch (150 mm) of the
natural soil is to be treated with lime for subgrade modification. The compaction of the
subgrade soil is specified to be to be 95% or greater of the standard Proctor density
with the moisture content equal to or not lower than 5% below the optimum moisture
content. The PCC slab thickness is 12.6 inch (320 mm). The 28-day design modulus of
rupture of concrete is 595 psi (4.1 MPa) and the 28-day elastic modulus is 4,003,042 psi
(27,600 MPa).
The full depth bituminous concrete pavement sections on both US-56 and I-70
have thickness of 13.4 inch (360 mm) and 18.1 inch (460 mm), respectively and are to
be constructed on a compacted subgrade (Figure 5.4). The detailed mixture properties
and binder specifications are listed in Table 5.1. The AASHTO soil classification of the
natural subgrade soil type is Non-Plastic A-3. The 6.0 inch (150 mm) of the natural soil
is to be treated with lime. The compaction specifications are the same as those for
JPCP.
The base year annual average daily traffic (AADT) on these sections is 18,200
on I-70 and 3,660 on US-56. The percent truck is 20% on I-70 and 5% on US-56.
5.5 Design Inputs of M-EPDG (Version 1.0) Analysis
Design inputs are the most significant aspects in the mechanistic-empirical
pavement design. The basic design inputs required for the M-EPDG analysis are: (1)
inputs for the pavement structures, such as, layer thickness, material type, etc.; (2)
68
inputs under which the pavement is designed to perform (traffic, climate, etc.); and
finally (3) inputs for the material and mix design properties of the layers.
The ME-Design Guide recommends that the designer use the subgrade resilient
modulus (MR) obtained from laboratory testing following AASHTO T 307 or NCHRP
Project 1-28. The following k k k1 2 3
consecutive model is used to predict the MR
value at the optimum moisture content (OMC);
2 3k k
octRopt 1 a
a a
M k p 1p p
Equation 5.1
Where,
RoptM Resilient modulus at OMC;
k ,k ,k1 2 3
Regression parameters;
ap Atmospheric pressure;
1 1 2 3J Stress invariant; and
2 2 2
oct 1 2 1 3 2 3
1
3Octahedral shear stress.
In the M-EPDG analysis, three levels of inputs can be provided to analyze and
design the pavement. “Level 1” is an advanced design procedure at the highest level of
reliability under heavy traffic condition. The generalized k k k1 2 3
model is directly
applied to ME-design if “Level 1” subgrade design input is selected. The design inputs
require site specific data collection and lab testing.
“Level 2” computes the subgrade resilient modulus using its relationship with
other subgrade properties, such as, the California Bearing Ratio (CBR) or R-value,
69
Dynamic Cone Penetration (DCP), water content and plasticity index (PI) (21). This data
is user specified and is used to routine design when lab data is not available.
For input “Level 3”, the resilient modulus at optimum moisture content is selected
on the basis of soil classification and sieve analysis. M-EPDG analysis also adjusts the
subgrade modulus for each design period (month) by accounting for the seasonal
variation in unbound material properties. The designer is allowed to provide modulus
value either for each design period or at optimum moisture content.
In this study, “Level 3” input with direct input of subgrade modulus was selected
to evaluate the subgrade modulus. The following section will briefly describe and
summarize all the inputs used in the M-EPDG analysis for the full depth asphalt
concrete (AC) and JPCP test sections.
5.5.1 General Inputs
In general inputs, “General Information” allows the designer to specify the
pavement type (flexible and rigid pavement), design life, pavement construction month,
traffic opening month etc. “Site Identification” provides information of a particular project
location, project ID, begin and end of mile posts and traffic direction (NB, SB, EB, and
WB). Analysis Parameters describes the analysis type and performance criteria to
predict the pavement performance over design life. The simulation of uncertainties and
variability occur during pavement design establish the reliability level by considering
certain potential errors of selected design inputs. Design reliability can be expressed as:
R = P [Distress over Design Period < Critical Distress Level] Equation 5.2
70
The parameters required in this design input section are IRI and the performance
criteria to verify the trial design. The designs that meet the applicable performance
criteria are then considered feasible from both structural and functional viewpoints. The
M-EPDG recommended reliability level of each distress type is based on the roadway
functional classification (1). In this study, the test sections were functionally classified as
Rural Interstate and Principal Arterials (rural). A design reliability of 90% was used for
all test sections based on the M-EPDG recommendations (24). Table 5.2 shows the
failure criteria for JPCP and AC pavement at 90% reliability level considered during
analysis.
71
Pavement
Type Performance Criteria Limit
Reliability
(%)
JPCP
Terminal IRI, (in/mile) 172 90
Transverse Crack (% slab crack) 15 90
Mean Joint Faulting, (inch) 0.12 90
Full Depth
Asphalt
Concrete
Initial IRI, (in/mile) 63 90
Terminal IRI, (in/mile) 172 90
AC surface down cracking, (ft/500) 2,000 90
AC bottom up cracking, (%) 25 90
AC thermal fracture, (ft/mile) 1,000 90
Chemically stabilized layer, (Fatigue Fracture) 25 90
Permanent Deformation, AC layer, (in) 0.25 90
Permanent Deformation, Total pavement, (in) 0.75 90
5.5.2 Design Traffic Inputs
Traffic data is a key data element for the design and analysis of pavement
structures in M-EPDG. The axle load distribution obtained from traffic module accurately
determines the axle loads that will be applied on the pavement during each time
increment of the damage accumulation process. To create this axle load distribution, the
software requires input data regarding the traffic volumes and loads, and hourly
distribution for the base year, and estimated growth over the design life.
The initial 2-way Annual Average Daily Traffic (AADT) are 18,200 and 3,660 for I-
70 and US-56, respectively. The percent heavy trucks are 20% and 5% for I-70 and US-
56, respectively. Both are divided roads with two lanes in each direction. The initial two
TABLE 5.2: Performance Criteria for JPC Pavement and Full Depth Asphalt Concrete
Pavement
72
way AADTT value was calculated using the percent heavy trucks of ADT and AADT
value obtained from the project traffic survey report. The percentages of truck in the
design lane and in the design direction are 100% and 60%, respectively on all test
locations. The posted speed limit on all test sections is 70 mph which was used as the
operational traffic speed for the base year in M-EPDG analysis.
Default values were used for other traffic volume adjustment factors, and monthly
adjustment factors in the M-EPDG analysis. A monthly adjustment factor of 1.0 was
used for all vehicle classes (4-13 by FHWA). The default values of vehicle class
distribution and hourly truck distributions were also used. A compound traffic growth
rate of 2% was used on both US-56 and I-70 project traffic survey report. For all test
sections, the default values recommended by M-EPDG level 3 traffic inputs for axle load
distribution factors, number of axle per truck and axle configurations were taken into
account during analysis.
5.5.3 Climate Inputs
The Enhanced Integrated Climatic Model (EICM) software is used to model
temperatures and moisture profiles in the pavement and subgrade. M-EPDG
recommends the user to accumulate the weather data from the weather stations in the
vicinity. It also suggests collecting the weather data for at least 24 months (25). In this
analysis, the project-specific virtual climatic data was generated by interpolating the
weather database from selected weather stations near the project area. The information
about the depth of ground water table was not available at the project locations. Hence,
M-EPDG recommended value of 10 feet was used for the depth of the ground water
table in the analysis.
73
5.5.4 Structural Inputs
This is the fourth set of inputs required by the M-E PDG software for evaluation
of the trial design. The inputs specify the structural, design, and material aspects of the
trial design chosen for the performance evaluation.
5.5.4.1 US-56 Full Depth Asphalt Concrete Pavement Section
According to the project investigation report, the top surface or wearing course of
US-56 pavement section has a 1.575-inch (40-mm) bituminous layer (SM-9.5T) with PG
70-28. The upper part of the base course or asphalt binder has a 2.36-inch (60-mm)
bituminous layer (SM-19A) with PG 70-28 and the lower asphalt base has a 9.45-inch
(240-mm) bituminous layer (SM-19A) with PG 64-22. Table 5.3 shows the input values
for the asphalt concrete layers in this project. Six inch (150 mm) “Natural Subgrade”
was considered in the M-EPDG run to investigate the sensitivity of the design towards
the compacted subgrade modulus.
5.5.4.2 I-70 Full Depth Asphalt Concrete and PCC Pavement Sections
Project reconstruction report on I-70 anticipates that the alternate 1 on pavement
section is a full depth asphalt concrete pavement (Figure 5.4). The top surface or
wearing course has a 1.575-inch (40 mm) bituminous layer (SM-9.5T) with PG 70-28.
The upper layer or the binder course has a 2.36-inch (60 mm) bituminous layer (SM-
19A) with PG 70-28 and the lower asphalt base has a 14.2-inch (360 mm) bituminous
layer (SM-19A) with PG 64-22. Table 5.3 shows the input parameters for the asphalt
concrete layers in the M-EPDG analysis. Six inch (150 mm) “Natural Subgrade” was
considered in the M-EPDG run.
74
Alternate 2 on I-70 section is a JPCP structure with stabilized base and
compacted subgrade. In this study, the project was assumed to have liquid sealant.
Dowel diameter was calculated from the PCC slab thickness as one-eighth of the slab
thickness in inches as shown in Table 5.4. No shoulder edge support was considered
during analysis. The selected width of the slab was the same of the lane width 12 ft
(3.66 m). The PCC mixture material and strength properties of the PCC slab are shown
in Table 5.4. The base is 4.0 inch (100 mm) thick and is chemically stabilized with
Portland cement. The base material properties, and layer strength used in the analysis
are also listed Table 5.4.
75
Input Parameters Input Values
I-70 US-56
Layer 1:(SM-9.5T)
Material Type
Layer Thickness, (inch)
Asphalt General:
Reference Temperature (0F)
Poisson’s Ratio
Effective Binder Content (%)
% Air Voids
Total Unit Weight (pcf)
Binder Grade (PG)
Asphalt Concrete
1.575
68
0.25
11.2
4
145.64
70-28
Asphalt Concrete
1.575
68
0.25
11.2
4
145.64
70-28
Layer 2:(SM-19A)
Material Type
Layer Thickness, (inch)
Asphalt General:
Reference Temperature (0F)
Poisson’s Ratio
Effective Binder Content (%)
% Air Voids
Total Unit Weight (pcf)
Binder Grade (PG)
Asphalt Concrete
2.36
68
0.25
10.5
4
146.52
70-28
Asphalt Concrete
2.36
68
0.25
10.5
4
146.52
70-28
Layer 3:(SM-19A)
Material Type
Layer Thickness, (inch)
Asphalt General:
Reference Temperature (0F)
Poisson’s Ratio
Effective Binder Content (%)
% Air Voids
Total Unit Weight (pcf)
Binder Grade (PG)
Asphalt Concrete
14.2
68
0.25
10.5
4
146.52
64-22
Asphalt Concrete
9.45
68
0.25
10.5
4
146.52
64-22
TABLE 5.3: Layer and Material Inputs for Flexible Pavement Analysis on I-
70 and US-56
76
*t = PCC slab thickness
“Natural Subgrade” of 6.0 inch (150 mm) was used in this analysis. The subgrade
moduli used were obtained from three different sources: (1) average stiffness obtained
by the IC roller on the test sections; (2) average backcalculated moduli obtained from
the Falling Weight Deflectometer tests at 10 different locations on each test section. The
backcalculation was done using an elastic layer backcalculation program and
Boussinesq‟s equation; and (3) from the laboratory resilient modulus testing of the
subgrade soil samples. Table 5.5 shows the input details for the subgrade layer. It is
interesting to note that the laboratory resilient modulus of the A-3 soil on US-56 is 3,481
psi (24 MPa) as compared to the 16,500 psi (113.8 MPa) recommended by M-EPDG for
A-3 soils. The FWD backcalculated and IC roller moduli are also lower than the
Input Parameters Input Values
Design Features: Joint Spacing, (ft) Dowel Bar Diameter, (t*/8), (inch) Dowel Bar Spacing, (inch)
15
1.575 12
Layer 1:PCC Slab Slab Thickness, (inch) Unit Weight, (pcf) Poisson‟s Ratio Mix Properties: Cement Type Cementitious Material Content, (lb/yd3) Water/Cement Ratio Aggregate Type Strength Properties: 28 Days Modulus of Rupture (psi) 28 Days Modulus of Elasticity (psi)
12.6 140 0.2
II
653.4 0.44
Limestone
595 4006,042
Layer 2: Cement Stabilized Base Layer Thickness, (inch) Unit Weight, (pcf) Poisson‟s Ratio Elastic/Resilient Modulus, (psi)
3.94 135 0.15
500,000
TABLE 5.4: Input Parameters for M-EPDG Rigid Pavement (JPCP) Analysis on I-70 Test
Section
77
recommended value. Although the backcalculated moduli from the FWD data are
somewhat closer, the discrepancy is more evident for the I-70 project. Here again for
the A-3 soil for the JPCP project, the M-EPDG recommended subgrade modulus value
is 16,000 psi (110.3 MPa). The subgrade modulus from the laboratory test is almost one
third of the M-EPDG value, and the moduli obtained from the FWD and IC roller are also
about one-third to about one-fourth of the recommended value. Although the in-situ
moisture contents on the I-70 test section during IC roller and FWD measurements
varied from +1% to +5% of the optimum moisture content, the difference is staggering.
78
TABLE 5.5: Subgrade Properties for Flexible/Rigid Pavement Analysis of I-70 and US-56
Test Sections
Input Parameters Input Values
I-70 US-56 “proof” US-56 “growth”
Soil Type Layer Thickness, (inch)
Poisson‟s Ratio
A-3 6.0
0.35
A-3 6.0
0.35
A-3 6.0
0.35
Input Subgrade Modulus FWD (Backcalculated), (psi) FWD (Boussinesq equation),
(psi) IC Roller, (psi) Lab Data, ((psi)
M-EPDG Recommended (A-3soil), (psi)
4,321 4,221 5,874 3,626
12,546 10,631 8,992 3,481
11,560 10,501 5,279 3,481
14,000 – 35,500
Plasticity Index (PI) Maximum Dry Density (MDD), (pcf)
Optimum Moisture Content (OMC), (%)
NP 114.7 11.7
NP 119.6 10.4
NP 125.6 10.2
79
5.6 Prediction on Distresses from M-EPDG Analysis
5.6.1 International Roughness Index (IRI)
The predicted IRI by the design analysis on both flexible and JPCP sections
passed the performance criteria assigned during the trial analysis. Table 5.6
summarizes the M-EPDG smoothness prediction after 20 years. In general, the table
shows that the IRI values increase with decreasing subgrade strength. The JPCP is less
sensitive to the subgrade modulus compared to the flexible pavements. On the I-70
section, the IRI value on the AC alternate section increases 4.3% with 38% decrease in
subgrade modulus while the increase in IRI for the JPCP alternate is only 0.9%. Similar
trend is evident on both the US-56 “proof” (IRI increases 9.9% with 72% decrease in
subgrade modulus) and the “growth” (IRI increases 9.2% with 69.8% decrease in
subgrade modulus) sections. However, the 20-year predicted IRI values indicate that
the effect of the subgrade modulus on this parameter is insignificant. This indicates that
the pavement designs obtained from the 1993 AASHTO Design Guide are sufficient
from a functional view point.
80
Project
Location
Pavement
Type
Subgrade
MR
(psi)
Predicted
Distress,
IRI
(inch/mile)
Predicted
Reliability
(%)
Target
Reliability
(%)
Comment
US-56
“proof” Flexible
12,546* 95.9 99.74 90 Pass
10,631** 96.7 99.7 90 Pass
8,992*** 97.6 99.65 90 Pass
3,481† 105.4 98.83 90 Pass
US-56
“growth” Flexible
11,560* 97 99.69 90 Pass
10,501** 97.5 99.65 90 Pass
5,279*** 102.6 99.21 90 Pass
3,481† 105.9 98.75 90 Pass
I-70
Flexible
5,874*** 109.9 97.93 90 Pass
4,321* 112.7 97.19 90 Pass
4,221** 113 97.13 90 Pass
3,626† 114.6 96.62 90 Pass
JPCP
5,874*** 74.5 99.99 90 Pass
4,321* 74.9 99.99 90 Pass
4,221** 74.9 99.99 90 Pass
3,626† 75.2 99.99 90 Pass
*FWD backcalculated modulus
** FWD Boussinesq equation
***IC Roller
†Lab Data
TABLE 5.6: IRI Evaluations by M-EPDG Analysis
81
5.6.2 Total Deformation
It is accepted that the total surface deformation is highly influenced by the
subgrade modulus. In fact, the currently used Shell and Asphalt Institute mechanistic-
empirical pavement design procedures use vertical compressive strain on the top of the
subgrade layer as one of the critical responses in design. The predicted distress of total
deformation by the M-EPDG analysis is also a significant output for the flexible
pavements. Figure 5.5 shows the total deformation on US-56 and I-70 sections.
The predicted total deformation decreases with increasing subgrade modulus for
all test sections. The total deformation on the US-56 sections varies from 0.22 inch (5.6
mm) to 0.47 inch (11.9 mm) while the I-70 flexible alternate section experiences
deformation of 0.51 inch (12.95 mm) to 0.63 inch (16 mm). Figure 5.5 (c) shows that the
I-70 flexible section failed at 90% reliability level due to decrease in subgrade modulus
(38%). The section also failed in the “deformation of the AC layer only” distress.
However, on US-56 sections, no failure was observed despite a 72% decrease in the
subgrade modulus though these sections are about 6 inches thinner than the I-70
asphalt section.
82
FIGURE 5.5: Predicted total deformation on (a) US-56 “proof” and (b) US-
56 “growth” and (c) I-70 section
88
90
92
94
96
98
100
102
3000 5000 7000 9000 11000 13000
Resilient Mudulus, MR, (psi)
Dis
tres
s R
elia
bil
ity,
(%
)
0.2
0.25
0.3
0.35
0.4
0.45
0.5
Pav
emen
t T
ota
l
Def
orm
atio
n,
(in
ch)
Predicted_Reliability Target_Reliability Total Deformation
88
90
92
94
96
98
100
102
3000 6000 9000 12000
Resilient Mudulus, MR, (psi)
Dis
tres
s R
elia
bil
ity,
(%
)
0.2
0.25
0.3
0.35
0.4
0.45
0.5
Pav
emen
t T
ota
l
Def
orm
atio
n,
(in
ch)
Predicted_Reliability Target_Reliability Total Deformation
80
90
100
110
3500 4000 4500 5000 5500 6000
Subgrade Modulus, MR, (psi)
Dis
tres
s R
elia
bil
ity,
(%
)
0.5
0.52
0.54
0.56
0.58
0.6
0.62
0.64P
avem
ent
To
tal
Def
orm
atio
n,
(in
ch)
Predicted_Reliability Target_Reliability Total Deformation
(a)
(b)
(c)
83
The possible reason could be the difference in AADTT values during the design
period on these project locations. The base year AADT on I-70 section is 18,200 with
20% truck traffic while US-56 has a base year AADT of 3,660 with 5% truck traffic. It
appears that the truck traffic repetitions at least early in the life of the pavement have a
very large influence on the permanent deformation of the flexible pavements. This also
raises the issue of the influence of the base thickness. A previous M-EPDG global
sensitivity analysis showed that the annual average daily truck traffic, asphalt concrete
(AC) thickness and subgrade strength have potential influence on flexible pavement
performance while effect of binder grade is insignificant (26). The effect of AC base was
studied here to understand the interaction between the AC base thickness and the
subgrade modulus.
The flexible pavements in this study were analyzed with respect to two different
AC base thicknesses. Figure 5.6 shows the variation in total pavement deformation and
the corresponding distress reliability for different subgrade moduli and AC base
thickness.
84
The figure illustrates that the total deformation increases rapidly with decreasing
AC base thickness. The difference in predicted deformation for two different thicknesses
remains the same irrespective of subgrade modulus. The difference is about 26% when
the subgrade modulus is equal to 3,626 psi (25 MPa). At a subgrade modulus of 5,874
psi (62 MPa), the difference is approximately 23%. Figure 5.6 also shows that for about
14.2 inch (360 mm) base thickness and subgrade modulus of 5,874 psi (62 MPa) the
total deformation performance criterion is satisfied with a predicted distress reliability of
94.3%. At base thickness of 9.45 inch (240 mm), the same subgrade modulus failed to
satisfy the same total deformation criterion (predicted distress reliability 60.4%). This
indicates that the AC base thickness has more influence on the total pavement
deformation than the foundation layer.
FIGURE 5.6: Effect of subgrade strength on total deformation at different
base thickness
0.5
0.6
0.7
0.8
0.9
1
3500 4000 4500 5000 5500 6000
Subgrade Modulus, MR, (psi)
To
tal
Defo
rmati
on
,
(in
ch
)
30
60
90
120
Dis
tress
Reli
ab
ilit
y,
(%)
T = 360 mm T = 240 mm
Predicted Reliability_T=360mm Predicted Reliability_T=240mm
Target Reliability
85
5.6.3 Effect of Subgrade Modulus on Design PCC Slab Thickness
The effect of subgrade modulus on the PCC slab thickness of JPCP pavements
was also investigated for different subgrade moduli. The slab thickness was varied from
9.5 inch (240 mm) to 13.5 inch (340 mm). The subgrade modulus was varied from 3,000
psi to 7,000 psi (48.3 MPa). Two key distresses on JPCP, IRI and percent slabs
cracked, were analyzed based on the preselected performance criteria. Figure 5.7
illustrates the results obtained from the design analysis.
86
FIGURE 5.7: Evaluation of (a) IRI and (b) % slabs cracked at different slab
thickness based on subgrade modulus
70
90
110
130
150
170
9 10 11 12 13 14
Slab Thickness, (inch)
Term
inal
IRI,
(in
ch
/mil
e)
60
90
120
Dis
tress
Reli
ab
ilit
y,
(%)
MR = 7000 (psi) MR = 6000 (psi)
MR = 5000 (psi) MR = 4000 (psi)
MR = 3000 (psi) Average Predicted Reliability
Target Reliability
(a)
(b)
0
1020
30
40
5060
70
9 10 11 12 13 14
Slab Thickness, (inch)
Perc
en
t S
lab
Cra
cked
, (%
)
0
20
40
60
80
100
120
Dis
tress
Reli
ab
ilit
y,
(%)
MR = 7000 (psi) MR = 6000 (psi)
MR = 5000 (psi) MR = 4000 (psi)
MR = 3000 (psi) Average Predicted Reliability
Target Reliability
87
The results in Figure 5.7 show that IRI and percent slabs cracked are not sensitive to
the subgrade modulus. However, both IRI and percent slabs cracked are significantly
influenced by the PCC slab thickness. The predicted distresses are satisfactory at 90%
target reliability up to the PCC slab thickness of 11 inch (279.4 mm). This may indicate
that the influence of subgrade modulus on slab thickness in the JPCP pavement design
is insignificant.
5.7 Determination of Target Modulus by M-EPDG Analysis
A study on intelligent compaction on Kansas highway embankments attempted to
develop “target” soil stiffness for pavement subgrade (27). Most European specifications
have developed target stiffness for soil subgrade depending upon soil type. The
success of the intelligent compaction control concept also fully depends on the
preassigened “target” modulus value of the compacted soil material. In this study, the
“target” modulus was taken as the minimum subgrade soil modulus that would be
required for a given base thickness and truck traffic while satisfying all performance
criteria at pre-selected design reliability (Figure 5.8).
88
Figure 5.8 shows the “target” subgrade stiffness values required on both project
locations at 90% reliability. It is evident that the “target” modulus decreases with
increasing base thickness on both test locations. On the US-56 section with a base
thickness of 9.45 inch (240 mm), the “target” subgrade stiffness value is about 2,290 psi
(15.8 MPa) and 4,737 psi (32.7 MPa), respectively, at different truck traffic volumes. On
FIGURE 5.8: “Target” subgrade modulus @ 90% reliability on (a) I-70 and
(b) US-56
1000
2000
3000
4000
5000
6000
7000
8.00 10.00 12.00 14.00 16.00
Base Thickness, (inch)
Targ
et
Mo
du
lus,
MR,
(psi)
AADTT = 183 AADTT = 1245
3000
5000
7000
9000
11000
13000
15000
17000
19000
21000
23000
8.00 9.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00
Base Thickness, (inch)
Targ
et
Mo
du
lus,
(psi)
(a)
(b)
89
the I-70 section, the “target” subgrade modulus value of 4,500 psi (31.0 MPa) is
obtained for the 14.2-inch (360 mm) base.
Test Section Method Subgrade Resilient Modulus, MR, (psi)
Measured/Calculated Target (M-EPDG)
US-56 “proof”
FWD
(Boussinesq) 10,524
2,290*
4,737** FWD (Backcalc) 12,438
IC Roller 8,449
Lab data 3,481
US-56 “growth”
FWD
(Boussinesq) 10,395
2,290*
4,737**
FWD
(Backcalc) 11,454
IC Roller 5,727
Lab data 3,481
I-70
FWD
(Boussinesq) 4,278
4,500*** FWD
(Backcalc) 4,179
IC Roller 5,674
Lab data 3,626
* AADTT = 183
** AADTT = 1,245
*** AADTT = 3640
Table 5.7 shows that the average moduli measured by the IC roller on US-56
“proof” section (8,449 psi), I-70 (5,874 psi) and on US-56 “growth” (5,727 psi) sections
are higher than the “target” values under different traffic conditions. The lab moduli that
were used in 1993 AASHTO Design Guide design for the US-56 sections are
satisfactory for low truck traffic condition (AADTT = 183) but not for higher AADTT
TABLE 5.7: “Target” Modulus and Measured/Calculated Modulus on US-56 and I-70
Sections
90
(1,245). Also on I-70, the “target” subgrade modulus is higher than the laboratory design
value. These results show that the “target” subgrade modulus for intelligent compaction
control can be derived based on the soil type and asphalt base thickness from the M-
EPDG analysis well in advance of construction.
5.8 Estimation of the Regression Constants for M-E Pavement Design
In M-E pavement design, the resilient modulus is estimated using a generalized
constitutive model (1). A linear or nonlinear regression analysis is used to fit the model
using laboratory generated resilient modulus test data. The nonlinear elastic coefficients
and exponents of the constitutive model are estimated using this nonlinear regression
analysis. The simplified general model (NCHRP 1-28A), derived from eq. 5.1 and used
in this analysis is as follows:
32kk
d 3 dr 1 a
a a
3 2M k p 1
p 3 p Equation 5.3
Where,
rM Resilient modulus, (psi);
1 Major principal stress (= d 3 );
2 Intermediate principal stress = 3 (for rM test on cylindrical specimen);
3 Minor principal stress/confining pressure;
d Deviator stress;
ap Normalizing stress (atmospheric pressure or 14.7 psi); and
1 2 3k ,k ,k Regression constants (obtained by fitting the laboratory Mr data to the
equation).
91
The simplified constitutive model was solved in this study using the SAS
nonlinear regression. From above relation, the coefficient k1 is proportional to the
Young‟s modulus and hence its values should be positive as Mr can not negative. A
higher value of bulk stress will result in higher Mr and the soil will behave as stiff
materials. Therefore, the regression exponent, k2, should be positive to support the
model output. Again, the exponent k3 is related to the octahedral shear stress and
should be negative since the increasing shear stress will result in softening of the
unbound material.
Estimation of nonlinear regression coefficients k1, k2 and k3 is extremely
important since they are used as level 1 input in M-E PDG design analysis to estimate
the actual Mr. As mentioned earlier, level 1 design procedure is highly recommended by
M-EPDG, when data is available, and is applicable to new pavement, reconstruction,
and rehabilitation design.
In this study, the resilient modulus testing was performed in the KDOT laboratory
using samples collected from three different test sections. The chamber confining
pressure ranged from 2 psi to 6 psi (13.8 kPa to 41.4 kPa) and the deviator stress
varied from 2 to 10 psi (13.8 to 68.9 kPa) for a particular confining pressure. The length
and diameter of the specimen were 5.52 inches (140 mm) and 2.85 inch (72 mm),
respectively. Moisture content ranged from OMC±2% at 90%, 95%, and 100%
compaction level. Table 5.8 shows the summary of test results on soil samples from
three different test sections.
92
Test
Section
%
Compaction
Moisture
Content
Confining
pressure,
3 ,
(psi)
Deviato
r Stress,
d , (psi)
Resilient
Modulus,
Mr,
(psi)
Comments
I-70
90
9.7 6 2 to 8 1,945 –
9,295
Failed,
partial Mr
11.7 - - - Failed
13.7 - - - Failed
95
9.7 6 2 to 6 4,942-5,480 Failed,
partial Mr
11.7 6 2 to 4 2,647-3,437 Failed,
partial Mr
13.7 - - - Failed
100
9.7 6 2 to 8 4,847-5,958 Failed,
partial Mr
11.7 6 2 to 6 4,367-5,141 Failed,
partial Mr
13.7 - - - Failed
US-56
“growth”
90
8.2 - - - Failed
10.2 6 2-10 5,200 Pass
12.2 6 2-6 3,260-4,334 Failed,
partial Mr
95
8.2 - - - Failed
10.2 6 2-10 3,589-4,151 Failed,
partial Mr
12.2 6 2-10 2,901-3,379 Failed,
partial Mr
100
8.2 - - - Failed
10.2 - - - Failed
12.2 6 2-6 2,851-3,231 Failed,
partial Mr
US-56
“proof”
90 8.3 - - - Failed
95 10.3 - - - Failed
100 12.3 - - - Failed
The estimated regression coefficients using the nonlinear SAS regression model
(Appendix A) are presented in Tables 5.9 and 5.10. No coefficients were obtained for
TABLE 5.8: Laboratory Resilient Modulus Test Results on US-56 and I-70 Sections
93
the US-56 “proof” section as the resilient modulus testing was unsuccessful on the
samples from that section.
%
Compaction
Moisture
Content k1 k2 k3
No. of
O
b
s
.
Comments
90
8.2 Fails 0
10.2 344 0.06 -0.097 15 Reasonable
12.2 648 -5.89 11.95 3 Unrealistic (k2<0 &
k3>0)
95
8.2 Fails 0
10.2 340 -1.55 2.82 6 Unrealistic (k2<0 &
k3>0)
12.2 220 0.32 -0.98 10 Reasonable
100
8.2 Fails 0
10.2 Fails 0
12.2 0.0142 45.38 -71.55 4
Unrealistic (k2 & k3
significantly high and
k1 very small)
TABLE 5.9: Estimated Regression Coefficient k1, k2 and k3 on US-56 Section
94
The success of nonlinear regression analysis is completely dependent on the
amount of available data. From these tables, it is clear that the coefficient values with
higher number of observations are realistic and reasonable. Figure 5.9 shows the
comparison of data between the Mr-measured (MR) and the Mr-calculated (MRHAT) from the
nonlinear regression analysis on the US-56 test section for two different sets of
observations.
TABLE 5.10: Estimated Regression Coefficient k1, k2 and k3 on I-70 Section
%
Compaction
Moisture
Content
k1 k2 k3 No. of
O
b
s
.
Comments
90
9.7 2.66 25.42 -45.51 4
Unrealistic (k2 & k3
significantly high and
k1 very small)
11.7 Fails 0
13.7 Fails 0
95
9.7 0.034 44.15 -71.46 3
Unrealistic (k2 & k3
significantly high and
k1 very small)
11.7 Fails 0
13.7 Fails 0
100
9.7 198.7 2.07 -2.07 4 Relatively
Reasonable
11.7 1.78 26.10 -45.85 3
Unrealistic (k2 & k3
significantly high and
k1 very small)
13.7 Fails 0
95
These figures show that the laboratory measured Mr and estimated Mr from
regression analysis are close to each other when the number of observations was 15.
However, when the number of observations was 6, no such correspondence was
observed.
FIGURE 5.9: Fitness of Mr data in SAS Nonlinear regression (a) (No. of
Obs.=15), (b) (No. of Obs.=6)
(a)
(b)
96
CHAPTER 6 - CONCLUSIONS AND RECOMMENDATIONS
6.1 Conclusions
This study was conducted to investigate the compaction quality control of
highway embankment soil in Kansas using a new technology called Intelligent
Compaction Control. Subgrade soil of three different test sections from two
reconstruction projects on US-56 and I-70 were compacted using IC roller. Other spot
tests such as stiffness measurements by soil gage, in-situ deflection measurements by
FWD and LFWD and DCP test were also performed on the test locations. Some
laboratory testing was also performed to characterize the test section subgrade soil.
Finally, M-EPDG (Mechanistic Empirical Pavement Design) analysis was performed to
investigate the effect of subgrade soil strength on the pavement distress level and also
to fix a “target” modulus for the field compaction quality control.
Based on this present study, the following conclusions can be made:
Due to continuous measurements of the in-situ stiffness of the subgrade soil
under compaction, the intelligent compaction (IC) roller is able to identify the soft
spots with lower stiffness in the spatial direction.
The IC roller stiffness is somewhat sensitive to the field moisture content. In-situ
moisture content close to the optimum moisture content will result in higher roller
stiffness.
“Target” stiffness values need to be selected in intelligent compaction control
because both very high and very low maximum densities achieved during
compaction may yield lower IC roller stiffness.
97
No good correlation was observed among IC roller stiffness and other
stiffness/modulus values obtained from the soil stiffness gage, LFWD, FWD and
DCP test data.
Correlation between the soil stiffness gage stiffness and the LFWD moduli was
found to be significant for the US-56 section. However, no such correlation was
found on I-70.
The reason for poor correlation is the discrepancy arises from different testing
devices. The discrepancy seems to occur due to the fact that different
equipments were capturing response from different volumes of soil on the same
test section. The lowest space volume was measured by soil stiffness gage while
the maximum volumetric depth was obtained by IC roller. FWD and LFWD
covered the intermediate volumetric depth of influence.
M-EPDG analysis shows that the predicted total pavement deformation and
roughness are sensitive to the subgrade modulus for flexible pavements. In
Jointed Plain Concrete Pavements, the key distresses are insensitive to the
subgrade modulus.
AC base thickness has more influence on the total pavement deformation than
the foundation layer. However, truck traffic plays an even more significant role in
controlling this distress.
The influence of subgrade modulus on the slab thickness in the JPCP pavement
design is insignificant.
98
The “target” subgrade modulus for intelligent compaction control roller can be
derived based on the soil type and asphalt base thickness well before
construction by doing M-EPDG analysis.
Nonlinear Regression Coefficients of constitutive model in M-E design are
estimated from laboratory resilient modulus test data. The coefficient values are
inconsistent since the resilient modulus test was not successful most of the time
and hence resulted less number of observations.
6.2 Recommendations
Based on this study, the following recommendations are made:
The subgrade soil stiffness is very sensitive to in-situ moisture content and IC
roller has been proven to be a very effective tool to identify those soft spots. Thus
the IC roller can be used for “proof rolling”.
Plate loading test is known to have a very good correlation with the IC roller
stiffness. FWD tests can be performed on subgrade using a larger diameter plate
(about 30 inches or 762 mm) to develop a better correlation between the IC roller
stiffness and the soil moduli backcalculated from the FWD data.
99
REFERENCES
1. Part 1: Introduction. Guide for Mechanistic-Empirical Design of New and
Rehabilitated Pavement Structures. Final Report, NCHRP Project 1-37A.
Transportation Research Board, National Research Council, Washington, D.C.,
July 2007.
2. BOMAG. Heavy Equipment, Compaction Control and Documentation System
BTM. Brochure, BOMAG Schriftenreihe, Boppard, Germany.
3. NCHRP, Intelligent Soil Compaction System. Request for Proposal for NCHRP
Project 21-09, National Cooperative Highway Research Program, Transportation
Research Board, Washington, D.C., September 2005.
4. Horan, B., and T. Ferragut, Intelligent Compaction Strategic Plan. FHWA, U. S.
Department of Transportation, April 2005.
5. Petersen, J. S., S. A. Romanoschi, M. A. Onyango, and M., Hossain. Evaluation
of Prima Light Falling-Weight Deflectometer as Quality Control Tool for
Compaction of Fine Grained Soils. Paper No. 07-1326, Preperint CD-ROM of the
86th Transportation Research Board Annual Meeting, Transportation Research
Board, Washington, D.C., January, 2007.
100
6. Briaud, J. L. and J. Seo, Intelligent Compaction: Overview and Research Needs.
Draft Report by the Texas A &M University, 2003. Available at
http://www.webs1.uidaho.edu/bayomy/trb/afh60/Intcompaction_Briaud_Septeme
ber2004.pdf
7. Kloubert, H., Intelligent VARIOCONTROL rollers with integrated quality control
system for soil compaction – principles, measurement, applications. Presentation
made at the 83rd Annual Meeting of the Transportation Research Board,
Washington, D.C., 2004.
8. Briaud, J.L. and J. Seo. Intelligent Compaction: Overview and Research Needs.
Draft Report by the Texas A &M University, December 2003.
9. Sandstrom, A. and C. Pettersson. Intelligent systems for QA/QC in soil
compaction, Presentation made at the 83rd Annual Meeting of the Transportation
Research Board, Washington, D.C., 2004.
10. Anderegg, R. and K. Kaufman. Intelligent Compaction with Vibratory Rollers;
Feedback Control Systems in Automatic Compaction and Compaction Control.
Transportation Research Record; Journal of the Transportation Research Board
No. 1868, Transportation Research Board, Washington, D.C., 2004, 124-134.
101
11. BOMAG Variocontrol, Automatic optimization of compaction on single drum
rollers for soils and granular base compaction, Technical Literature, 2004.
12. Kim, D., T. Nantung, N. Z. Siddiki, and J. R. Kim. Implementation of the New
Mechanistic-Empirical Design for Subgrade Materials-Indiana Experience. TRB
Preprint CD-ROM Paper #07-0463, 86th Annual Meeting of the Transportation
Research Board, Washington, D.C., 2007.
13. Petersen, D.L., J. Siekmeier, C. Nelson, and R. L. Peterson. Intelligent Soil
Compaction Technology: Results and a Roadmap Toward Widespread Use.
Transportation Research Record: Journal of the Transportation Research Board,
No. 1975, Transportation Research Board, Washington, D.C., 2006, pp. 81-88.
14. Krober, W., R. Floss, and W. Wallrah. Dynamic Soil Stiffness as Quality Criterion
for Soil Compaction. Technical Paper, BOMAG Schriftenreihe, Boppard,
Germany.
15. Anderegg, R. ACE AMMANN Compaction Expart. Technical Paper, Ammann
Verdichtung AG, Langenthal, Swiss.
102
16. Sawangsuriya, A., J. P. Bosscher, and B. T. Edil, Laboratory Evaluation of the
Soil Stiffness Gauge. In Transportation Research Record: Journal of the
Transportation Research Board, No 1808, TRB, National Research Council,
Washington, D. C., 2002, pp. 30-37.
17. George, K.P. Falling Weight Deflectometer for Estimating Subgrade Resilient
Moduli. Final Report No. FHWA/MS-DOT-RD-03-153, Mississippi Department of
Transportation, Jackson, December 2003.
18. Livneh, M., and I. Ishai. Pavement and Material Evaluation by a Dynamic Cone
Penetrometer. Proc., Sixth International Conference on the Structural Design of
Asphalt Pavement, Vol. 1, Ann Arbor, Michigan, 1987, pp. 665-674.
19. Mau, R. R. Correlation of the Automated Cone Penetrometer, Dynamic Cone
Penetrometer, and California Bearing Ratio. M.S. Thesis. Department of Civil
Engineering, University of Kansas, Lawrence, January 1989.
20. Webster, S. L., R. H. Grau, and T. P. Williams. Description and Application of
Dual Mass Dynamic Cone Penetrometer. USACE Waterways Experiment
Station, Vicksburg, Mississippi, May 1992.
103
21. Part 2: Design Inputs. Guide for Mechanistic-Empirical Design of New and
Rehabilitated Pavement Structures. Final Report, NCHRP Project 1-37A.
Transportation Research Board, National Research Council, Washington, D.C.,
July 2007.
22. SAS. Statistical Analysis System, The SAS Institute, Carey, North Carolina.,
1999
23. Petersen, D. L., M. L. Erickson, R. Roberson, and J. Siekmeier. Intelligent Soil
Compaction: Geostatistical Data Analysis and Construction Specifications. Paper
No.07-2858, Preperint CD-ROM of the 86th Transportation Research Board
Annual Meeting, Transportation Research Board, Washington, D.C., January,
2007.
24. Khazanovich, L., C. Celauro, B. Chadbourn, J. Zollars, and S. Dai. Evaluation of
Subgrade Resilient Modulus Predictive Model for Use in Mechanistic-Empirical
Pavement Design Guide. Transportation Research Record: Journal of the
Transportation Research Board, n 1947, Transportation Research Board of the
National Academics, Washington, D.C., 2006, p 155-166.
25. Barry, C. R. and C. Schwartz. Geotechnical Aspects of Pavements. Report no.
FHWA NHI-05-037. National Highway Institute, Federal Highway Administration,
Washington, D.C., January 2005.
104
26. Graves, R. C., and K. C. Mahboub. Pilot Study in Sampling-Based Sensitivity
Analysis of NCHRP Design Guide for Flexible Pavements. Transportation
Research Record: Journal of the Transportation Research Board, n 1947,
Transportation Research Board of the National Academics, Washington, D.C.,
2006, p 123-135.
27. Rahman, F., M. Hossain, M. M. Hunt, and S. A. Romanoschi. Intelligent
Compaction Control of Highway Embankment Soil. TRB Preprint CD-ROM Paper
#07-2962, 86th Annual Meeting of the Transportation Research Board,
Washington, D.C., 2007.
105
APPENDIX A
SAS NONLINEAR REGRESSION ANALYSIS
INPUT AND OUTPUT FILES
106
SAS Input File:
US-56 Test section (90% Compaction @ OMC)
data Resilient;
input sigma3 sigmad MR;
datalines;
6 2 5131.3
6 4 5052.8
6 6 5004.1
6 8 5193.4
6 10 5185.5
4 2 5038.0
4 4 4837.6
4 6 4954.0
4 8 5235.7
4 10 5124.3
2 2 4968.9
2 4 4784.6
2 6 4924.6
2 8 5210.8
2 10 4490.5
;
proc nlin data=Resilient method=GAUSS noitprint hougaard;
parms k1=0
k2=0
k3=-6;
model MR=k1*14.7*(((sigmad+3*sigma3)/14.7)**k2)*(((0.4714*sigmad/14.7)+1)**k3);
der.k1= 14.7*(((sigmad+3*sigma3)/14.7)**k2)*(((0.4714*sigmad/14.7)+1)**k3);
der.k2=
(k1*14.7*(((sigmad+3*sigma3)/14.7)**k2)*(((0.4714*sigmad/14.7)+1)**k3))*log((
sigmad+3*sigma3)/14.7);
der.k3=
(k1*14.7*(((sigmad+3*sigma3)/14.7)**k2)*(((0.4714*sigmad/14.7)+1)**k3))*log((
0.4714*sigmad/14.7)+1);
output out=new p=MRhat r=MRresid
run;
proc plot data=new;
plot MR*sigmad='*' MRhat*sigmad='+' / overlay;
run;
107
SAS Output File:
US-56 Test section (90% Compaction @ OMC)
Estimation Summary
Method Gauss-Newton
Iterations 6
Subiterations 1
Average Subiterations 0.166667
R 1.415E-6
PPC(k3) 7.185E-6
RPC(k3) 0.000677
Object 1.873E-8
Objective 426059.9
Observations Read 15
Observations Used 15
Observations Missing 0
The NLIN Procedure
Sum of Mean Approx
Source DF Squares Square F Value Pr >F
Regression 3 3.7648E8 1.2549E8 3534.57 <.0001
Residual 12 426060 35505.0
Uncorrected Total 15 3.7691E8
Corrected Total 14 548638
Approx Approximate 95%
Parameter Estimate Std Error Confidence Limits Skewness
k1 343.5 8.6757 324.6 362.4 0.0421
k2 0.0600 0.0327 -0.0113 0.1313 0.00893
k3 -0.0976 0.1468 -0.4174 0.2221 0.00643
Approximate Correlation Matrix
k1 k2 k3
k1 1.0000000 0.2966818 -0.9057739
k2 0.2966818 1.0000000 -0.4981702
k3 -0.9057739 -0.4981702 1.0000000